{ "text": "You've probably typed something into Gemini , got an answer ,\nand closed the tab , the same way you'd use any other chatbot .\nHere's the thing, that's maybe 10% of what it actually does.\nUse it right , and it can quietly take over almost half the busy\nwork you're still doing by hand .\nI spent hours mapping every model , every mode ,\nand every product Gemini is quietly wired into ,\nand the number that stopped me was this .\nGemini's app alone has over 650 million monthly users ,\nand that's before you count everyone using it inside search .\nMost of them are using maybe 20% of it, with no idea the rest even\nexists .\nLook at this data from the Census Bureau .\nOnly about one in five US businesses actually use AI in their operations\n.\nSo , if you run a business , and you're even thinking about this\n, you're ahead of most of your competition .\nWhat you might not know is that alongside covering AI news ,\nwe work with business owners to help them implement AI in their\nbusiness .\nOur engineering team gets to know how your business runs ,\nthen builds the automation with you .\nYou'll find the link in the description below .\nClick it , fill out a short form about your business ,\nand we'll get in touch to set up a call .\nSo , in this video , I'm breaking down exactly what Gemini is as\nof mid 2026 , every current model , every mode ,\nand everywhere Google has quietly built it in .\nBy the end , you'll know exactly which Gemini tool to reach for\ndepending on what you're actually trying to do ,\ninstead of just typing into whichever box is in front of you .\nFirst , let's clear up the biggest misconception .\nGemini isn't one product at all .\nWhat Gemini actually is , here's the mental model you need before\nany of this makes sense .\nGemini isn't a single AI , it's Google's umbrella name for a whole\nplatform , a family of models underneath , and a set of products\non top that let you actually talk to them .\nThink of it in two layers .\nThe bottom layer is the models themselves, things like Gemini\n3 .\n6 Flash or Gemini 3 .\n1 Pro .\nThese are the engines tuned for different jobs .\nSome built for speed , some for heavy reasoning ,\nsome for images or audio .\nYou never see these names unless you go looking .\nThe top layer is everything you actually click on .\nThe Gemini app , AI mode inside Google search ,\nGemini inside Gmail and Docs , the voice assistant on your phone\n.\nAll of those are just different doors into the same underlying\nmodels .\nThat's the whole point of this video .\nGoogle isn't trying to build one great chatbot .\nIt's trying to put the same AI brain behind every product you already\nuse .\nSo , let's start with the brains , the actual models ,\nbecause once you know what each one is built for ,\neverything else clicks into place .\nThe current model lineup .\nThis is a demo checklist , so we're going model by model .\nWhat it is , what it's actually good for , and where you can get\nit .\nGemini 3 .\n7 flash .\nLaunched on August 13th , 2026 , this is Google's newest flash\nmodel and its most capable workhorse yet .\nIt's built primarily for coding and AI agents with major improvements\nin software engineering , web development , and complex multi step\nworkflows .\nGoogle has already made it generally available through the Gemini\nAPI , positioning 3 .\n7 flash as the new go to model when you want strong intelligence\nwithout giving up the speed and efficiency the flash lineup is\nknown for .\nGemini 3 .\n6 flash .\nThis is Google's current flagship , announced in a company blog\npost on July 21st , 2026 .\nIt's built as a workhorse , strong at coding ,\nknowledge work , and multimodal tasks .\nAnd according to Google's own numbers , it does the job using about\n17 fewer tokens on average than its predecessor .\nFewer tokens means faster answers and a lower bill if you're paying\nfor it through the API .\nYou can reach it through the Gemini API , through AI Studio ,\nor simply by using the Gemini app and searches AI mode .\nNo extra setup required .\nIf you only remember one model name from this video ,\nmake it this one because it's what most of Gemini is quietly running\non right now .\nGemini 3 .\n5 Flash .\nThis one launched back in May 2026 and it's the model that was\nactually powering AI mode in search before 3 .\n6 Flash took over .\nGoogle described it as frontier level intelligence at exceptional\nspeed and by its own benchmarks , it pushed output throughput to\nroughly four times faster than other top models at the time .\nIt's still very much in active use across the Gemini app ,\nGoogle's anti gravity platform , and enterprise tools .\nIt's slightly behind 3 .\n6 now , but it's the model quietly sitting behind a huge share\nof what shipped this year .\nGemini 3 .\n5 Flashlight .\nSame July announcement , different job entirely .\nThis is the stripped down , high throughput sibling .\nGoogle sites roughly 350 tokens per second , which is built for\nvolume , not depth .\nYou wouldn't use this for a hard reasoning problem .\nYou'd use it for background tasks and agent pipelines that need\nto move fast and cheap at scale .\nGemini 3 .\n1 Pro .\nReleased in February 2026 , this is the reasoning specialist .\nGoogle's benchmarks claim roughly double the logic performance\nof the earlier Gemini 3 Pro .\nIt's mostly gated behind preview access through the API ,\nanti gravity , Vertex AI , and Pro or Ultra subscriptions in the\nconsumer app .\nIf 3 .\n6 Flash is built for speed , 3 .\n1 Pro is built for depth .\nThe model you'd want on a genuinely hard problem ,\nnot a quick one .\nAnd if you're actually building on top of these through the API\n, price is where the real world decision gets made .\nAccording to Google's own published rates , 3 .\n6 Flash runs about 1 .\n50 per million input tokens and 7 .\n50 per million output tokens .\nFor context , that's noticeably cheaper on the output side than\nGPT 5 .\n6 Luna's roughly 6 per million .\nThat gap is exactly why so many developers default to flash tier\nmodels for anything running at volume .\nNow , a handful of specialty models worth knowing by name ,\neven if we don't dwell on each one .\nNano Banana 2 is Gemini's current image generation and editing\nmodel , replacing the older Imagen line entirely .\nAnd that's not a small detail because Imagen is actually shutting\ndown on August 17th , 2026 .\nIf you've had workflows built on Imagen , that clock is already\nrunning .\nThere's also a Nano Banana 2 light variant built purely for speed\n, trading a small amount of quality for much faster ,\ncheaper output at high volume .\nVIO 3 .\n1 is Google's video generation model , still in beta ,\nbuilt to turn a text prompt into a short clip with matching audio\n.\nGemini audio 3 .\n5 live translate handles real time speech to speech translation\nacross more than 70 languages , already built into Google Meet\nand Android .\nAnd if you're curious about the more niche end of the lineup ,\nthere's a security focused variant called 3 .\n5 flash cyber built to coordinate with vulnerability scanning tools\n.\nAnd Lyra 3 .\n5 , Google's music model , which can now generate tracks up to\n3 minutes long from a text prompt .\nHere's the honest limitation worth naming .\nGoogle ships a lot of these models fast , and the naming gets confusing\non purpose or not .\n3.5,\n3.6,\n3.1 Pro,\nFlash, Flash Cyber.\nIf you're not building on top of the API professionally ,\nyou genuinely don't need to memorize this list .\nYou just need to know the shape of it .\nFast and cheap , deep reasoning , and multimodal .\nThat's really three categories wearing a lot of different name\ntags .\nThe modes you actually interact with , models are the engine .\nModes are the steering wheel .\nHere's where things get useful for anyone who isn't a developer\n.\nAI mode inside Google Search turns your search bar into a conversation\n.\nAs of I/O 2026, it's globally powered by Gemini 3.5 Flash and instead of 10 blue links, you get a written answer\n5 flash and instead of 10 blue links , you get a written answer\nwith follow up questions and sometimes an interactive widget built\non the fly .\nAnyone with search can use it .\nNo subscription required .\nAsk something like , What's a quick dinner with what's in my fridge\n?\nand it answers in full sentences , not a list of recipe blogs .\nDeep Think is the extra effort version of the Gemini app .\nIt spends more compute per answer to reason through harder problems\nstep by step .\nGoogle gates this one behind Google AI Ultra and it's built for\ngenuinely difficult science or engineering questions ,\nnot everyday chat .\nNow , Deep Research is where this stops being a chatbot and starts\nbeing an assistant .\nYou give it a topic and instead of one reply ,\nit plans a research strategy , opens web pages ,\nreads them and if you allow it , pulls from your own Gmail and\nDrive , too .\nWhat comes back isn't a paragraph .\nIt's a full multi page report inside Gemini's canvas .\nThis is the part of Gemini that actually earns the word agent and\nwe're coming back to why that matters in a few minutes .\nGemini Live is the voice and camera mode .\nSay , Hey Google , let's chat and you're talking to it hands free\nwith the option to point your camera at something and ask what\nit's looking at live .\nAnd Canvas is the workspace mode .\nType , Create a quiz app about planets and it writes the code ,\nthe interface and the content in one pass , right there for you\nto edit .\nOne more worth a mention briefly because it's still early .\nGemini Spark , a personal agent announced at IO 2026 ,\nmeant to run continuously in the background handling things like\nscheduling .\nRight now , it's limited to early Ultra testers .\nSo , treat this one as coming , not here .\nQuick gut check before we move on .\nIf all of that sounds like a lot of separate tools ,\nthat's fair .\nBut , notice the pattern .\nEvery single one of these modes is just Gemini 3 .\n5 or 3 .\n6 flash wearing a different job title .\nYou're not learning six different AIs , you're learning six different\nways to ask the same brain for help .\nMultimodal in practice .\nLet's talk about what multimodal actually means day to day ,\nbecause it's more than a buzzword on a slide .\nGemini reads and writes text and code .\nObviously , that's the baseline .\nBut drop a photo into a chat and ask it to caption or edit it ,\nand Nano Banana handles that .\nAsk it to speak an answer out loud , and Gemini's audio models\ngenerate that voice on the spot with actual control over tone and\npacing .\nAsk for a short video and VO builds one from scratch .\nAsk it to edit an existing clip , swap the sky ,\nchange the style , and that's a separate tool called Gemini Omni\ndoing frame by frame editing by voice command .\nInside Google Docs and Sheets , the same underlying models can\ndraft a document from your meeting notes or build a spreadsheet\nout of a pile of invoices , complete with formulas and charts .\nNot just raw numbers dumped into cells .\nIn Slides , hand it a list of bullet points and it can lay out\nan actual deck , not just text on blank slides .\nAnd through Gemini Live's camera mode , you can point your phone\nat a menu in a language you don't speak and get a live translation\noverlaid on what you're looking at .\nOr ask it to identify an object it's looking at through the lens\n.\nNo typing involved .\nHere's the part worth remembering .\nYou never pick the model .\nYou just say what you want .\nMake this an infographic .\nTranslate this .\nWrite this in Python .\nAnd Gemini quietly roots the request to whichever model actually\ndoes that job .\nThat's the design philosophy in one sentence .\nOne platform , and it decides the plumbing so you don't have to\n.\nWhere Gemini actually lives .\nThis is the part that's easy to underestimate .\nGemini isn't confined to one app .\nIt's spread across nearly everything Google ships .\nIn Search , it's AI mode , already covered .\nIn Gmail , it's behind Smart Compose and auto reply suggestions\n.\nIn Docs , Sheets and Slides , Ultra and Pro Pro Pro get Gemini\ndrafting text , building formulas , and designing slide layouts\n, pulling context from your own files when you let it .\nIn Drive , it can find and summarize documents for you .\nIn Google Meet , it's doing live caption translation .\nOn Android , especially Pixel devices , it's baked straight into\nthe voice assistant .\nAnd there's a Chrome extension that lets the browser send page\ncontent straight to Gemini , so you can ask questions about whatever\ntab you're on .\nAnd for developers , all of it is exposed through Google AI Studio\nand the Gemini API , plus a newer platform called Antigravity for\nbuilding multi agent workflows on top of it .\nThe strategic point here isn't subtle .\nGoogle isn't trying to win the best standalone chatbot argument\n.\nIt's trying to make sure you're never more than one product away\nfrom Gemini , no matter what you're doing on a Google device or\nin a Google app .\nAgents : the part that actually matters .\nNow , here's the shift I promised earlier , the one that actually\nchanges what this platform is for .\nEverything so far has been ask a question , get an answer .\nAgents are Google trying to move Gemini past that entirely .\nDeep research is the clearest example already live ,\nplan , browse , synthesize , write , without you babysitting every\nstep .\nSpark is the early , still limited attempt at a persistent personal\nagent running continuously in the background .\nAnd on the developer side , Antigravity lets companies build coordinated\nteams of sub agents .\nGoogle's own blog post gave an example of businesses running parallel\nagents to analyze data at scale , rather than one model doing everything\nsequentially .\nPicture the difference in practice .\nThe old way , you ask Gemini , What should I know before a trip\nto Japan ?\nand it gives you a paragraph .\nThe agent way , you say , Plan my trip to Japan .\nand it checks flights , compares hotel options ,\nand comes back with an actual itinerary , pausing to confirm with\nyou before it books anything .\nThat's the same underlying model , just given permission to take\nmore than one step before handing control back to you .\nNone of this is science fiction anymore , and none of it is fully\nfinished , either .\nThat's the honest read .\nDeep Research genuinely works today .\nSpark is still in early testing , but the direction is unmistakable.\n.\nbreak it into steps , and execute most of them without you typing\na follow up for every single one .\nWhat actually makes Gemini different ?\nSo , how does this stack up against everyone else building the\nsame kind of thing ?\nLet's be balanced here , because Google's advantages are real ,\nbut so are its weak spots .\nThe clearest edge is data .\nGemini can pull from live search results , Maps ,\nGmail , and Drive in ways that a closed sandbox chatbot simply\ncan't match without plugins bolted on .\nThe second edge is reach .\nEvery Android phone is a potential Gemini client ,\nand every Workspace business account already has it available .\nNo competitor has that kind of built in distribution .\nAnd on raw benchmarks , Gemini 3 Pro topped the LM Arena leaderboard\n, which , regardless of how much weight you put on any single leaderboard\n, says Google's infrastructure and DeepMind's research are producing\nreal , top tier results , not just hype .\nBut , and this matters for credibility , Google is genuinely more\nconservative about rollout than some competitors .\nDeep Think and Spark are still gated behind ultra subscriptions\nor limited testing , while some rivals ship new capabilities to\neveryone at once .\nAnd the tier structure itself , free , pro , ultra ,\nAPI pricing , can be genuinely confusing next to a simpler flat\nsubscription from a competitor .\nIf you've ever opened the Gemini pricing page and closed it 5 minutes\nlater still unsure which plan you need , that's not just you .\nWhere this is actually headed , a few things are confirmed and\na few are still rumor , and it's worth keeping those separate .\nConfirmed , Gemini 3 .\n5 Pro is currently in partner testing with a public release expected\nsoon , and Google has already started training on Gemini 4 ,\nsoon , and Google has already started training on Gemini 4 ,\naccording to its own July 2026 announcement, though there's no\npublic timeline for that yet.\nWorkspace AI rollout continues expanding, and Gemini Live's regional\nlanguage support keeps growing.\nSpeculative and worth labeling clearly as such,\nthere's talk of a dedicated on-device AI chip for future Pixel\nphones, and some experimental DeepMind research around 3D avatars\nand world simulation that hasn't shipped as a product.\nTreat both of those as possible, not coming.\nNothing official has confirmed either one.\nThe verdict.\nSo, where does that leave things?\nGemini in 2026 isn't a chatbot you occasionally open.\nIt's an AI layer Google has threaded through search,\nGmail, your documents, and increasingly your phone itself.\nThe models handle the thinking, the modes handle how you ask,\nand agents like Deep Research are the clearest sign of where all\nof it is actually heading.\nIf there's one thing worth trying this week, it's Deep Research\non something you'd normally spend an evening looking into yourself\n, and actually watching it work instead of just reading the final\nreport.\nDrop a comment with which piece of this surprised you most,\nthe model lineup, the agent side, or just how much of this you\nwere already using without realizing it.\nI'll be back soon with a deeper breakdown on how Deep Research\nactually performs against a real research task.\nThanks for watching, and I'll see you in the next one.", "language": "en", "duration": 997.248, "segments": [ { "start": 0.0, "end": 3.01, "text": "You've probably typed something into Gemini , got an answer ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 3.02, "end": 6.552, "text": "and closed the tab , the same way you'd use any other chatbot .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 6.562, "end": 10.372, "text": "Here's the thing, that's maybe 10% of what it actually does.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Here's the thing , that's maybe 10 of what it actually does .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 10.382, "end": 13.835, "text": "Use it right , and it can quietly take over almost half the busy", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 13.845, "end": 15.666, "text": "work you're still doing by hand .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 15.676, "end": 18.849, "text": "I spent hours mapping every model , every mode ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 18.859, "end": 21.954, "text": "and every product Gemini is quietly wired into ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 21.964, "end": 23.983, "text": "and the number that stopped me was this .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 23.993, "end": 28.315, "text": "Gemini's app alone has over 650 million monthly users ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 28.325, "end": 31.52, "text": "and that's before you count everyone using it inside search .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 31.53, "end": 35.496, "text": "Most of them are using maybe 20% of it, with no idea the rest even", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Most of them are using maybe 20 of it , with no idea the rest even", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 35.506, "end": 36.469, "text": "exists .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 36.479, "end": 39.025, "text": "Look at this data from the Census Bureau .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 39.035, "end": 43.527, "text": "Only about one in five US businesses actually use AI in their operations", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 43.537, "end": 44.392, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 44.402, "end": 47.313, "text": "So , if you run a business , and you're even thinking about this", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 47.323, "end": 50.394, "text": ", you're ahead of most of your competition .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 50.404, "end": 53.876, "text": "What you might not know is that alongside covering AI news ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 53.886, "end": 56.477, "text": "we work with business owners to help them implement AI in their", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 56.487, "end": 57.958, "text": "business .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 57.968, "end": 60.879, "text": "Our engineering team gets to know how your business runs ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 60.889, "end": 63.08, "text": "then builds the automation with you .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 63.09, "end": 65.48, "text": "You'll find the link in the description below .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 65.49, "end": 68.4, "text": "Click it , fill out a short form about your business ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 68.41, "end": 70.119, "text": "and we'll get in touch to set up a call .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 70.129, "end": 74.279, "text": "So , in this video , I'm breaking down exactly what Gemini is as", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 74.289, "end": 78.719, "text": "of mid 2026 , every current model , every mode ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 78.729, "end": 81.518, "text": "and everywhere Google has quietly built it in .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 81.528, "end": 85.118, "text": "By the end , you'll know exactly which Gemini tool to reach for", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 85.128, "end": 87.478, "text": "depending on what you're actually trying to do ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 87.488, "end": 90.557, "text": "instead of just typing into whichever box is in front of you .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 90.567, "end": 93.237, "text": "First , let's clear up the biggest misconception .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 93.247, "end": 95.477, "text": "Gemini isn't one product at all .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 95.487, "end": 99.277, "text": "What Gemini actually is , here's the mental model you need before", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 99.287, "end": 100.717, "text": "any of this makes sense .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 100.727, "end": 104.476, "text": "Gemini isn't a single AI , it's Google's umbrella name for a whole", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 104.486, "end": 108.076, "text": "platform , a family of models underneath , and a set of products", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 108.086, "end": 110.236, "text": "on top that let you actually talk to them .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 110.246, "end": 111.795, "text": "Think of it in two layers .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 111.805, "end": 114.835, "text": "The bottom layer is the models themselves, things like Gemini", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "The bottom layer is the models themselves , things like Gemini", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 114.845, "end": 115.368, "text": "3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 115.378, "end": 117.315, "text": "6 Flash or Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 117.325, "end": 118.195, "text": "1 Pro .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 118.205, "end": 120.915, "text": "These are the engines tuned for different jobs .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 120.925, "end": 123.794, "text": "Some built for speed , some for heavy reasoning ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 123.804, "end": 125.634, "text": "some for images or audio .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 125.644, "end": 127.994, "text": "You never see these names unless you go looking .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 128.004, "end": 130.434, "text": "The top layer is everything you actually click on .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 130.444, "end": 133.673, "text": "The Gemini app , AI mode inside Google search ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 133.683, "end": 137.473, "text": "Gemini inside Gmail and Docs , the voice assistant on your phone", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 137.483, "end": 137.793, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 137.803, "end": 140.673, "text": "All of those are just different doors into the same underlying", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 140.683, "end": 141.433, "text": "models .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 141.443, "end": 143.072, "text": "That's the whole point of this video .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 143.082, "end": 145.752, "text": "Google isn't trying to build one great chatbot .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 145.762, "end": 149.152, "text": "It's trying to put the same AI brain behind every product you already", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 149.162, "end": 149.832, "text": "use .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 149.842, "end": 153.112, "text": "So , let's start with the brains , the actual models ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 153.122, "end": 155.751, "text": "because once you know what each one is built for ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 155.761, "end": 157.671, "text": "everything else clicks into place .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 157.681, "end": 159.231, "text": "The current model lineup .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 159.241, "end": 162.631, "text": "This is a demo checklist , so we're going model by model .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 162.641, "end": 165.99, "text": "What it is , what it's actually good for , and where you can get", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 166.0, "end": 166.39, "text": "it .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 166.4, "end": 167.35, "text": "Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 167.36, "end": 168.83, "text": "7 flash .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 168.84, "end": 172.595, "text": "Launched on August 13th , 2026 , this is Google's newest flash", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 172.605, "end": 175.119, "text": "model and its most capable workhorse yet .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 175.129, "end": 178.885, "text": "It's built primarily for coding and AI agents with major improvements", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 178.895, "end": 182.69, "text": "in software engineering , web development , and complex multi step", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 182.7, "end": 183.892, "text": "workflows .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 183.902, "end": 186.616, "text": "Google has already made it generally available through the Gemini", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 186.626, "end": 188.84, "text": "API , positioning 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 188.85, "end": 192.746, "text": "7 flash as the new go to model when you want strong intelligence", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 192.756, "end": 195.549, "text": "without giving up the speed and efficiency the flash lineup is", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 195.559, "end": 196.27, "text": "known for .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 196.28, "end": 197.285, "text": "Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 197.295, "end": 198.354, "text": "6 flash .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 198.364, "end": 201.599, "text": "This is Google's current flagship , announced in a company blog", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 201.609, "end": 204.196, "text": "post on July 21st , 2026 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 204.206, "end": 206.831, "text": "It's built as a workhorse , strong at coding ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 206.841, "end": 209.305, "text": "knowledge work , and multimodal tasks .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 209.315, "end": 212.577, "text": "And according to Google's own numbers , it does the job using about", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 212.587, "end": 216.167, "text": "17 fewer tokens on average than its predecessor .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 216.177, "end": 219.919, "text": "Fewer tokens means faster answers and a lower bill if you're paying", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 219.929, "end": 221.515, "text": "for it through the API .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 221.525, "end": 225.465, "text": "You can reach it through the Gemini API , through AI Studio ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 225.475, "end": 229.267, "text": "or simply by using the Gemini app and searches AI mode .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 229.277, "end": 230.994, "text": "No extra setup required .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 231.004, "end": 233.685, "text": "If you only remember one model name from this video ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 233.695, "end": 237.139, "text": "make it this one because it's what most of Gemini is quietly running", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 237.149, "end": 238.02, "text": "on right now .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 238.03, "end": 239.061, "text": "Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 239.071, "end": 240.101, "text": "5 Flash .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 240.111, "end": 244.342, "text": "This one launched back in May 2026 and it's the model that was", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 244.352, "end": 247.71, "text": "actually powering AI mode in search before 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 247.72, "end": 249.104, "text": "6 Flash took over .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 249.114, "end": 252.465, "text": "Google described it as frontier level intelligence at exceptional", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 252.475, "end": 256.506, "text": "speed and by its own benchmarks , it pushed output throughput to", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 256.516, "end": 259.746, "text": "roughly four times faster than other top models at the time .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 259.756, "end": 263.068, "text": "It's still very much in active use across the Gemini app ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 263.078, "end": 266.629, "text": "Google's anti gravity platform , and enterprise tools .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 266.639, "end": 268.017, "text": "It's slightly behind 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 268.027, "end": 271.551, "text": "6 now , but it's the model quietly sitting behind a huge share", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 271.561, "end": 272.991, "text": "of what shipped this year .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 273.001, "end": 273.918, "text": "Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 273.928, "end": 275.072, "text": "5 Flashlight .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 275.082, "end": 278.073, "text": "Same July announcement , different job entirely .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 278.083, "end": 280.834, "text": "This is the stripped down , high throughput sibling .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 280.844, "end": 284.955, "text": "Google sites roughly 350 tokens per second , which is built for", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 284.965, "end": 286.476, "text": "volume , not depth .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 286.486, "end": 288.716, "text": "You wouldn't use this for a hard reasoning problem .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 288.726, "end": 292.198, "text": "You'd use it for background tasks and agent pipelines that need", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 292.208, "end": 294.198, "text": "to move fast and cheap at scale .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 294.208, "end": 295.092, "text": "Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 295.102, "end": 295.999, "text": "1 Pro .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 296.009, "end": 300.08, "text": "Released in February 2026 , this is the reasoning specialist .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 300.09, "end": 303.041, "text": "Google's benchmarks claim roughly double the logic performance", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 303.051, "end": 304.922, "text": "of the earlier Gemini 3 Pro .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 304.932, "end": 308.283, "text": "It's mostly gated behind preview access through the API ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 308.293, "end": 312.604, "text": "anti gravity , Vertex AI , and Pro or Ultra subscriptions in the", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 312.614, "end": 313.684, "text": "consumer app .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 313.694, "end": 314.325, "text": "If 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 314.335, "end": 316.525, "text": "6 Flash is built for speed , 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 316.535, "end": 318.566, "text": "1 Pro is built for depth .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 318.576, "end": 321.207, "text": "The model you'd want on a genuinely hard problem ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 321.217, "end": 322.247, "text": "not a quick one .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 322.257, "end": 324.928, "text": "And if you're actually building on top of these through the API", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 324.938, "end": 327.929, "text": ", price is where the real world decision gets made .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 327.939, "end": 330.57, "text": "According to Google's own published rates , 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 330.58, "end": 332.517, "text": "6 Flash runs about 1 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 332.527, "end": 335.185, "text": "50 per million input tokens and 7 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 335.195, "end": 337.092, "text": "50 per million output tokens .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 337.102, "end": 340.933, "text": "For context , that's noticeably cheaper on the output side than", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 340.943, "end": 342.467, "text": "GPT 5 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 342.477, "end": 344.975, "text": "6 Luna's roughly 6 per million .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 344.985, "end": 348.776, "text": "That gap is exactly why so many developers default to flash tier", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 348.786, "end": 351.176, "text": "models for anything running at volume .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 351.186, "end": 354.898, "text": "Now , a handful of specialty models worth knowing by name ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 354.908, "end": 356.778, "text": "even if we don't dwell on each one .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 356.788, "end": 360.7, "text": "Nano Banana 2 is Gemini's current image generation and editing", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 360.71, "end": 364.041, "text": "model , replacing the older Imagen line entirely .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 364.051, "end": 367.351, "text": "And that's not a small detail because Imagen is actually shutting", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 367.361, "end": 370.362, "text": "down on August 17th , 2026 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 370.372, "end": 373.522, "text": "If you've had workflows built on Imagen , that clock is already", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 373.532, "end": 374.181, "text": "running .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 374.191, "end": 377.694, "text": "There's also a Nano Banana 2 light variant built purely for speed", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 377.704, "end": 380.79, "text": ", trading a small amount of quality for much faster ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 380.8, "end": 382.753, "text": "cheaper output at high volume .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 382.763, "end": 383.602, "text": "VIO 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 383.612, "end": 386.864, "text": "1 is Google's video generation model , still in beta ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 386.874, "end": 389.975, "text": "built to turn a text prompt into a short clip with matching audio", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 389.985, "end": 390.312, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 390.322, "end": 391.97, "text": "Gemini audio 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 391.98, "end": 396.217, "text": "5 live translate handles real time speech to speech translation", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 396.227, "end": 399.783, "text": "across more than 70 languages , already built into Google Meet", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 399.793, "end": 400.655, "text": "and Android .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 400.665, "end": 403.468, "text": "And if you're curious about the more niche end of the lineup ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 403.478, "end": 406.137, "text": "there's a security focused variant called 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 406.147, "end": 410.185, "text": "5 flash cyber built to coordinate with vulnerability scanning tools", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 410.195, "end": 410.64, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 410.65, "end": 412.027, "text": "And Lyra 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 412.037, "end": 415.879, "text": "5 , Google's music model , which can now generate tracks up to", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 415.889, "end": 418.019, "text": "3 minutes long from a text prompt .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 418.029, "end": 420.238, "text": "Here's the honest limitation worth naming .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 420.248, "end": 424.041, "text": "Google ships a lot of these models fast , and the naming gets confusing", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 424.051, "end": 425.23, "text": "on purpose or not .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 425.24, "end": 425.765, "text": "3.5,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "3 .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 425.775, "end": 426.905, "text": "3.6,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "5 , 3 .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 426.915, "end": 427.98, "text": "3.1 Pro,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "6 , 3 .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 427.99, "end": 430.82, "text": "Flash, Flash Cyber.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "1 pro , flashlight , flash cyber .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 430.83, "end": 433.909, "text": "If you're not building on top of the API professionally ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 433.919, "end": 436.677, "text": "you genuinely don't need to memorize this list .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 436.687, "end": 438.522, "text": "You just need to know the shape of it .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 438.532, "end": 441.611, "text": "Fast and cheap , deep reasoning , and multimodal .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 441.621, "end": 444.298, "text": "That's really three categories wearing a lot of different name", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 444.308, "end": 445.181, "text": "tags .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 445.191, "end": 448.952, "text": "The modes you actually interact with , models are the engine .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 448.962, "end": 450.476, "text": "Modes are the steering wheel .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 450.486, "end": 453.324, "text": "Here's where things get useful for anyone who isn't a developer", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 453.334, "end": 453.765, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 453.775, "end": 457.656, "text": "AI mode inside Google Search turns your search bar into a conversation", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 457.666, "end": 458.218, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 458.228, "end": 462.544, "text": "As of I/O 2026, it's globally powered by Gemini 3.5 Flash and instead of 10 blue links, you get a written answer", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "As of IO 2026 , it's globally powered by Gemini 3 .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 462.554, "end": 466.361, "text": "5 flash and instead of 10 blue links , you get a written answer", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 466.371, "end": 469.771, "text": "with follow up questions and sometimes an interactive widget built", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 469.781, "end": 470.573, "text": "on the fly .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 470.583, "end": 472.178, "text": "Anyone with search can use it .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 472.188, "end": 473.782, "text": "No subscription required .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 473.792, "end": 476.706, "text": "Ask something like , What's a quick dinner with what's in my fridge", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 476.716, "end": 477.072, "text": "?", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 477.082, "end": 480.699, "text": "and it answers in full sentences , not a list of recipe blogs .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 480.709, "end": 484.005, "text": "Deep Think is the extra effort version of the Gemini app .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 484.015, "end": 487.773, "text": "It spends more compute per answer to reason through harder problems", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 487.783, "end": 488.815, "text": "step by step .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 488.825, "end": 492.382, "text": "Google gates this one behind Google AI Ultra and it's built for", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 492.392, "end": 495.468, "text": "genuinely difficult science or engineering questions ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 495.478, "end": 496.831, "text": "not everyday chat .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 496.841, "end": 500.558, "text": "Now , Deep Research is where this stops being a chatbot and starts", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 500.568, "end": 501.641, "text": "being an assistant .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 501.651, "end": 504.366, "text": "You give it a topic and instead of one reply ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 504.376, "end": 507.653, "text": "it plans a research strategy , opens web pages ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 507.663, "end": 511.701, "text": "reads them and if you allow it , pulls from your own Gmail and", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 511.711, "end": 512.622, "text": "Drive , too .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 512.632, "end": 514.585, "text": "What comes back isn't a paragraph .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 514.595, "end": 518.234, "text": "It's a full multi page report inside Gemini's canvas .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 518.244, "end": 522.362, "text": "This is the part of Gemini that actually earns the word agent and", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 522.372, "end": 525.047, "text": "we're coming back to why that matters in a few minutes .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 525.057, "end": 527.652, "text": "Gemini Live is the voice and camera mode .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 527.662, "end": 531.54, "text": "Say , Hey Google , let's chat and you're talking to it hands free", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 531.55, "end": 534.226, "text": "with the option to point your camera at something and ask what", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 534.236, "end": 535.829, "text": "it's looking at live .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 535.839, "end": 537.913, "text": "And Canvas is the workspace mode .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 537.923, "end": 541.761, "text": "Type , Create a quiz app about planets and it writes the code ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 541.771, "end": 545.448, "text": "the interface and the content in one pass , right there for you", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 545.458, "end": 546.089, "text": "to edit .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 546.099, "end": 549.416, "text": "One more worth a mention briefly because it's still early .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 549.426, "end": 554.145, "text": "Gemini Spark , a personal agent announced at IO 2026 ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 554.155, "end": 557.031, "text": "meant to run continuously in the background handling things like", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 557.041, "end": 557.953, "text": "scheduling .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 557.963, "end": 560.598, "text": "Right now , it's limited to early Ultra testers .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 560.608, "end": 563.043, "text": "So , treat this one as coming , not here .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 563.053, "end": 565.047, "text": "Quick gut check before we move on .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 565.057, "end": 567.532, "text": "If all of that sounds like a lot of separate tools ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 567.542, "end": 568.374, "text": "that's fair .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 568.384, "end": 569.696, "text": "But , notice the pattern .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 569.706, "end": 572.676, "text": "Every single one of these modes is just Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 572.686, "end": 573.664, "text": "5 or 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 573.674, "end": 575.909, "text": "6 flash wearing a different job title .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 575.919, "end": 579.316, "text": "You're not learning six different AIs , you're learning six different", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 579.326, "end": 581.52, "text": "ways to ask the same brain for help .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 581.53, "end": 583.192, "text": "Multimodal in practice .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 583.202, "end": 586.257, "text": "Let's talk about what multimodal actually means day to day ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 586.267, "end": 588.527, "text": "because it's more than a buzzword on a slide .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 588.537, "end": 590.756, "text": "Gemini reads and writes text and code .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 590.766, "end": 592.707, "text": "Obviously , that's the baseline .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 592.717, "end": 596.17, "text": "But drop a photo into a chat and ask it to caption or edit it ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 596.18, "end": 597.842, "text": "and Nano Banana handles that .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 597.852, "end": 601.284, "text": "Ask it to speak an answer out loud , and Gemini's audio models", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 601.294, "end": 605.089, "text": "generate that voice on the spot with actual control over tone and", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 605.099, "end": 605.81, "text": "pacing .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 605.82, "end": 609.135, "text": "Ask for a short video and VO builds one from scratch .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 609.145, "end": 612.259, "text": "Ask it to edit an existing clip , swap the sky ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 612.269, "end": 616.024, "text": "change the style , and that's a separate tool called Gemini Omni", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 616.034, "end": 618.668, "text": "doing frame by frame editing by voice command .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 618.678, "end": 622.393, "text": "Inside Google Docs and Sheets , the same underlying models can", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 622.403, "end": 625.838, "text": "draft a document from your meeting notes or build a spreadsheet", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 625.848, "end": 629.843, "text": "out of a pile of invoices , complete with formulas and charts .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 629.853, "end": 632.166, "text": "Not just raw numbers dumped into cells .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 632.176, "end": 635.611, "text": "In Slides , hand it a list of bullet points and it can lay out", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 635.621, "end": 638.936, "text": "an actual deck , not just text on blank slides .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 638.946, "end": 642.02, "text": "And through Gemini Live's camera mode , you can point your phone", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 642.03, "end": 645.479, "text": "at a menu in a language you don't speak and get a live translation", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 645.489, "end": 647.149, "text": "overlaid on what you're looking at .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 647.159, "end": 650.229, "text": "Or ask it to identify an object it's looking at through the lens", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 650.239, "end": 650.607, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 650.617, "end": 651.919, "text": "No typing involved .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 651.929, "end": 653.588, "text": "Here's the part worth remembering .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 653.598, "end": 655.059, "text": "You never pick the model .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 655.069, "end": 656.41, "text": "You just say what you want .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 656.42, "end": 658.0, "text": "Make this an infographic .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 658.01, "end": 659.233, "text": "Translate this .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 659.243, "end": 660.624, "text": "Write this in Python .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 660.634, "end": 664.026, "text": "And Gemini quietly roots the request to whichever model actually", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 664.036, "end": 665.188, "text": "does that job .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 665.198, "end": 667.754, "text": "That's the design philosophy in one sentence .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 667.764, "end": 671.141, "text": "One platform , and it decides the plumbing so you don't have to", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 671.151, "end": 671.401, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 671.411, "end": 673.165, "text": "Where Gemini actually lives .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 673.175, "end": 675.57, "text": "This is the part that's easy to underestimate .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 675.58, "end": 677.935, "text": "Gemini isn't confined to one app .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 677.945, "end": 680.621, "text": "It's spread across nearly everything Google ships .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 680.631, "end": 683.467, "text": "In Search , it's AI mode , already covered .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 683.477, "end": 686.794, "text": "In Gmail , it's behind Smart Compose and auto reply suggestions", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 686.804, "end": 687.315, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 687.325, "end": 691.355, "text": "In Docs , Sheets and Slides , Ultra and Pro Pro Pro get Gemini", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 691.365, "end": 695.418, "text": "drafting text , building formulas , and designing slide layouts", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 695.428, "end": 698.454, "text": ", pulling context from your own files when you let it .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 698.464, "end": 701.884, "text": "In Drive , it can find and summarize documents for you .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 701.894, "end": 704.755, "text": "In Google Meet , it's doing live caption translation .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 704.765, "end": 708.345, "text": "On Android , especially Pixel devices , it's baked straight into", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 708.355, "end": 709.501, "text": "the voice assistant .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 709.511, "end": 712.213, "text": "And there's a Chrome extension that lets the browser send page", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 712.223, "end": 715.683, "text": "content straight to Gemini , so you can ask questions about whatever", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 715.693, "end": 716.64, "text": "tab you're on .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 716.65, "end": 720.508, "text": "And for developers , all of it is exposed through Google AI Studio", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 720.518, "end": 724.56, "text": "and the Gemini API , plus a newer platform called Antigravity for", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "and the Gemini API , plus a newer platform called antigravity for", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 724.57, "end": 727.008, "text": "building multi agent workflows on top of it .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 727.018, "end": 729.175, "text": "The strategic point here isn't subtle .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 729.185, "end": 732.084, "text": "Google isn't trying to win the best standalone chatbot argument", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 732.094, "end": 732.425, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 732.435, "end": 735.073, "text": "It's trying to make sure you're never more than one product away", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 735.083, "end": 738.403, "text": "from Gemini , no matter what you're doing on a Google device or", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 738.413, "end": 739.526, "text": "in a Google app .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 739.536, "end": 742.215, "text": "Agents : the part that actually matters .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Agents , the part that actually matters .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 742.225, "end": 745.224, "text": "Now , here's the shift I promised earlier , the one that actually", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 745.234, "end": 747.35, "text": "changes what this platform is for .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 747.36, "end": 751.323, "text": "Everything so far has been ask a question , get an answer .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 751.333, "end": 754.693, "text": "Agents are Google trying to move Gemini past that entirely .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 754.703, "end": 757.863, "text": "Deep research is the clearest example already live ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 757.873, "end": 762.076, "text": "plan , browse , synthesize , write , without you babysitting every", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 762.086, "end": 762.757, "text": "step .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 762.767, "end": 766.489, "text": "Spark is the early , still limited attempt at a persistent personal", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 766.499, "end": 769.178, "text": "agent running continuously in the background .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 769.188, "end": 772.98, "text": "And on the developer side , Antigravity lets companies build coordinated", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "And on the developer side , antigravity lets companies build coordinated", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 772.99, "end": 774.42, "text": "teams of sub agents .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 774.43, "end": 777.859, "text": "Google's own blog post gave an example of businesses running parallel", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 777.869, "end": 782.02, "text": "agents to analyze data at scale , rather than one model doing everything", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 782.03, "end": 782.98, "text": "sequentially .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 782.99, "end": 784.82, "text": "Picture the difference in practice .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 784.83, "end": 788.38, "text": "The old way , you ask Gemini , What should I know before a trip", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 788.39, "end": 789.26, "text": "to Japan ?", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 789.27, "end": 790.66, "text": "and it gives you a paragraph .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 790.67, "end": 794.06, "text": "The agent way , you say , Plan my trip to Japan .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 794.07, "end": 796.98, "text": "and it checks flights , compares hotel options ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 796.99, "end": 800.34, "text": "and comes back with an actual itinerary , pausing to confirm with", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 800.35, "end": 801.9, "text": "you before it books anything .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 801.91, "end": 805.02, "text": "That's the same underlying model , just given permission to take", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 805.03, "end": 807.98, "text": "more than one step before handing control back to you .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 807.99, "end": 810.98, "text": "None of this is science fiction anymore , and none of it is fully", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 810.99, "end": 811.98, "text": "finished , either .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 811.99, "end": 813.22, "text": "That's the honest read .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 813.23, "end": 815.58, "text": "Deep Research genuinely works today .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 815.59, "end": 819.08, "text": "Spark is still in early testing , but the direction is unmistakable.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Spark is still in early testing , but the direction is unmistakable", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 819.09, "end": 819.62, "text": ".", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 819.63, "end": 822.42, "text": "break it into steps , and execute most of them without you typing", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Google wants Gemini to eventually take a task ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 822.43, "end": 826.02, "text": "a follow up for every single one .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "break it into steps , and execute most of them without you typing", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 826.03, "end": 828.02, "text": "What actually makes Gemini different ?", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "a follow up for every single one .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 828.03, "end": 830.14, "text": "So , how does this stack up against everyone else building the", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "What actually makes Gemini different ?", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 830.15, "end": 832.781, "text": "same kind of thing ?", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "So , how does this stack up against everyone else building the", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 832.791, "end": 834.061, "text": "Let's be balanced here , because Google's advantages are real ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "same kind of thing ?", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 834.071, "end": 837.542, "text": "but so are its weak spots .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Let's be balanced here , because Google's advantages are real ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 837.552, "end": 839.183, "text": "The clearest edge is data .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "but so are its weak spots .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 839.193, "end": 840.824, "text": "Gemini can pull from live search results , Maps ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "The clearest edge is data .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 840.834, "end": 843.825, "text": "Gmail , and Drive in ways that a closed sandbox chatbot simply", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Gemini can pull from live search results , Maps ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 843.835, "end": 848.587, "text": "can't match without plugins bolted on .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Gmail , and Drive in ways that a closed sandbox chatbot simply", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 848.597, "end": 850.868, "text": "The second edge is reach .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "can't match without plugins bolted on .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 850.878, "end": 852.468, "text": "Every Android phone is a potential Gemini client ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "The second edge is reach .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 852.478, "end": 855.47, "text": "and every Workspace business account already has it available .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Every Android phone is a potential Gemini client ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 855.48, "end": 858.951, "text": "No competitor has that kind of built in distribution .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "and every Workspace business account already has it available .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 858.961, "end": 862.032, "text": "And on raw benchmarks , Gemini 3 Pro topped the LM Arena leaderboard", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "No competitor has that kind of built in distribution .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 862.042, "end": 866.034, "text": ", which , regardless of how much weight you put on any single leaderboard", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "And on raw benchmarks , Gemini 3 Pro topped the LM Arena leaderboard", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 866.044, "end": 869.815, "text": ", says Google's infrastructure and DeepMind's research are producing", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": ", which , regardless of how much weight you put on any single leaderboard", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 869.825, "end": 873.717, "text": "real , top tier results , not just hype .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": ", says Google's infrastructure and DeepMind's research are producing", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 873.727, "end": 876.598, "text": "But , and this matters for credibility , Google is genuinely more", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "real , top tier results , not just hype .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 876.608, "end": 880.359, "text": "conservative about rollout than some competitors .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "But , and this matters for credibility , Google is genuinely more", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 880.369, "end": 882.867, "text": "Deep Think and Spark are still gated behind ultra subscriptions", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "conservative about rollout than some competitors .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 882.877, "end": 885.96, "text": "or limited testing , while some rivals ship new capabilities to", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Deep Think and Spark are still gated behind ultra subscriptions", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 885.97, "end": 889.16, "text": "everyone at once .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "or limited testing , while some rivals ship new capabilities to", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 889.17, "end": 890.36, "text": "And the tier structure itself , free , pro , ultra ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "everyone at once .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 890.37, "end": 894.84, "text": "API pricing , can be genuinely confusing next to a simpler flat", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "And the tier structure itself , free , pro , ultra ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 894.85, "end": 898.799, "text": "subscription from a competitor .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "API pricing , can be genuinely confusing next to a simpler flat", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 898.809, "end": 900.44, "text": "If you've ever opened the Gemini pricing page and closed it 5 minutes", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "subscription from a competitor .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 900.45, "end": 903.525, "text": "later still unsure which plan you need , that's not just you .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "If you've ever opened the Gemini pricing page and closed it 5 minutes", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 903.535, "end": 907.029, "text": "Where this is actually headed , a few things are confirmed and", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "later still unsure which plan you need , that's not just you .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 907.039, "end": 909.93, "text": "a few are still rumor , and it's worth keeping those separate .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Where this is actually headed , a few things are confirmed and", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 909.94, "end": 912.87, "text": "Confirmed , Gemini 3 .", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "a few are still rumor , and it's worth keeping those separate .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 912.88, "end": 914.69, "text": "5 Pro is currently in partner testing with a public release expected", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Confirmed , Gemini 3 .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 914.7, "end": 918.067, "text": "soon , and Google has already started training on Gemini 4 ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "5 Pro is currently in partner testing with a public release expected", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 918.077, "end": 921.532, "text": "soon , and Google has already started training on Gemini 4 ,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 921.542, "end": 924.795, "text": "according to its own July 2026 announcement, though there's no", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "according to its own July 2026 announcement , though there's no", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 924.805, "end": 926.406, "text": "public timeline for that yet.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "public timeline for that yet .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 926.416, "end": 930.475, "text": "Workspace AI rollout continues expanding, and Gemini Live's regional", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Workspace AI rollout continues expanding , and Gemini Live's regional", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 930.485, "end": 932.127, "text": "language support keeps growing.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "language support keeps growing .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 932.137, "end": 935.309, "text": "Speculative and worth labeling clearly as such,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Speculative and worth labeling clearly as such ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 935.319, "end": 938.613, "text": "there's talk of a dedicated on-device AI chip for future Pixel", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "there's talk of a dedicated on device AI chip for future Pixel", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 938.623, "end": 942.529, "text": "phones, and some experimental DeepMind research around 3D avatars", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "phones , and some experimental DeepMind research around 3D avatars", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 942.539, "end": 945.597, "text": "and world simulation that hasn't shipped as a product.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "and world simulation that hasn't shipped as a product .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 945.607, "end": 948.785, "text": "Treat both of those as possible, not coming.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Treat both of those as possible , not coming .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 948.795, "end": 950.976, "text": "Nothing official has confirmed either one.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Nothing official has confirmed either one .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 950.986, "end": 951.773, "text": "The verdict.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "The verdict .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 951.783, "end": 953.367, "text": "So, where does that leave things?", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "So , where does that leave things ?", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 953.377, "end": 957.033, "text": "Gemini in 2026 isn't a chatbot you occasionally open.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Gemini in 2026 isn't a chatbot you occasionally open .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 957.043, "end": 959.982, "text": "It's an AI layer Google has threaded through search,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "It's an AI layer Google has threaded through search ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 959.992, "end": 964.285, "text": "Gmail, your documents, and increasingly your phone itself.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Gmail , your documents , and increasingly your phone itself .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 964.295, "end": 967.752, "text": "The models handle the thinking, the modes handle how you ask,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "The models handle the thinking , the modes handle how you ask ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 967.762, "end": 970.661, "text": "and agents like Deep Research are the clearest sign of where all", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 970.671, "end": 972.055, "text": "of it is actually heading.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "of it is actually heading .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 972.065, "end": 975.124, "text": "If there's one thing worth trying this week, it's Deep Research", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "If there's one thing worth trying this week , it's Deep Research", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 975.134, "end": 977.773, "text": "on something you'd normally spend an evening looking into yourself", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 977.783, "end": 980.822, "text": ", and actually watching it work instead of just reading the final", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 980.832, "end": 981.539, "text": "report.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "report .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 981.549, "end": 984.328, "text": "Drop a comment with which piece of this surprised you most,", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Drop a comment with which piece of this surprised you most ,", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 984.338, "end": 987.715, "text": "the model lineup, the agent side, or just how much of this you", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "the model lineup , the agent side , or just how much of this you", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 987.725, "end": 989.628, "text": "were already using without realizing it.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "were already using without realizing it .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 989.638, "end": 992.417, "text": "I'll be back soon with a deeper breakdown on how Deep Research", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow" }, { "start": 992.427, "end": 995.127, "text": "actually performs against a real research task.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "actually performs against a real research task .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "start": 995.137, "end": 999.201, "text": "Thanks for watching, and I'll see you in the next one.", "source": "source_caption", "timing_source": "vad_whisper_monotonic_reflow", "text_original": "Thanks for watching , and I'll see you in the next one .", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" } ], "transcript_selection": { "schema": "videonote.transcript_selection.v1", "status": "SELECTED", "requested_language": "en", "duration": 997.248, "reference_path": "/Users/sagawa/AI_WORK/video_notes/url/20260817_164301__Google_Gemini_Explained_80_of_Google_Gemini_You_Never_Use_2026_Guide/reference_subtitle.vtt", "reference_kind": "automatic_original", "reference_language": "en-orig", "selected_source": "youtube_auto_original", "decision_reason": "source_timing_retained_after_quality_gate", "whisper_rejected": false, "source_profile": { "segment_count": 454, "multilingual": false, "first_start": 0.0, "final_end": 999.2, "duration_coverage": 1.0, "maximum_gap_seconds": 0.0, "inverted_count": 0, "language": { "requested": "en", "matched": true, "latin": 13376, "han": 0, "kana": 0, "hangul": 0 }, "hallucination_region_count": 0, "hallucination_regions": [], "usable_full": true, "usable_partial": true }, "whisper_profile": { "segment_count": 231, "multilingual": false, "first_start": 0.0, "final_end": 997.12, "duration_coverage": 0.999872, "maximum_gap_seconds": 1.04, "inverted_count": 0, "language": { "requested": "en", "matched": true, "latin": 13378, "han": 0, "kana": 0, "hangul": 0 }, "hallucination_region_count": 0, "hallucination_regions": [], "usable_full": true, "usable_partial": true }, "comparison": { "window_seconds": 30.0, "compared_window_count": 34, "median_score": 0.9521, "divergent_window_count": 0, "divergent_ratio": 0.0, "divergent_windows": [] }, "timestamp_alignment": { "applied": false, "estimated_offset_seconds": 0.0, "matched_segment_count": 0, "match_scores": [] }, "caption_timing_repair": { "method": "monotonic_multi_anchor_piecewise_interpolation", "applied": false, "anchor_count": 0, "anchors": [], "source_block_count": 86, "whisper_block_count": 106, "matched_block_count": 76, "case_normalization": { "applied": false, "changed_segment_count": 0 }, "source_text_preserved": true, "reason": "source_timing_retained_after_quality_gate" }, "filled_source_profile": { "segment_count": 87, "multilingual": false, "first_start": 0.0, "final_end": 999.2, "duration_coverage": 1.0, "maximum_gap_seconds": 0.0, "inverted_count": 0, "language": { "requested": "en", "matched": true, "latin": 13376, "han": 0, "kana": 0, "hangul": 0 }, "hallucination_region_count": 0, "hallucination_regions": [], "usable_full": true, "usable_partial": true }, "whisper_duration_guard": { "media_duration": 997.248, "rejected_segment_count": 0, "rejected_segments": [] } }, "caption_alignment": { "method": "vad_whisper_monotonic_reflow", "applied": true, "source_text_authority": "preserved" }, "presegmentation_correction": { "applied": true, "correction_count": 70 } }