{
  "text": "Quen 3.8 Max just dropped.\nI've already plugged it into Hermes Agent and it's pretty powerful from what I've tested so far.\nSo, what I'm going to show you today is how to use Quen 3.8 Max with Hermes Agent and also how to get the most out of it and some of the cool stuff that we've created with it so far.\nSo, the great thing about Quen 3.8 Max is that it's basically Fable 5 level and you can use your coding plan with it to plug it in directly into Hermes, which you actually can't do with Fable 5 unless you use the API.\nIt's also a 2.4 trillion parameter brain and it actually beats Fable 5 on some of the key benchmarks.\nSo basically Hermes agent is the hands and then you can change the brain.\nSo you can plug in Quen 3.8 Max.\nThis was a 2.4 trillion parameters, 1 million context window and also designed for agentic tasks.\nI'll show you some things we built with it so far.\nSo, it created like some pretty cool coding apps as you can see right here.\nBut also, we actually linked it to MCP.\nSo, we linked it to Hicksfield MCP just to test it out and see what it could do.\nAnd we actually used and gave it our book image as an example to see if it could code and build some cool stuff with Higsfield MCP and Quen 3.8 Max inside Hermes agent.\nSo, this is some of the stuff that he created, which is pretty nice and it was super easy to use.\nThe other thing I would say about Quen 3.8 Max, it's pretty fast to respond.\nHere's another example.\nSo, what we actually have inside our Hermes agentic operating system is all of our different profiles with different models plugged in.\nNow, one thing I'll say about Quen 3.0 Max is it's very fast to apply.\nSo, we say, for example, learn and then give it access to a guide.\nAnd you can see here it says skill learned and registered.\nHere's a skill.\nHere's the path that saved locally as a skill MD file, which we can open up in a second.\nAnd here's what it captures.\nAnd so we've taught it a new skill in the space of a couple of minutes.\nAnd if we open this up, we've got the full skill that's written here based on the guide we actually trained it with.\nNow, one thing I definitely recommend with this is that when you're using new models, whether that's Fable 5 or whether that's Quinn 3.8 Max or whatever comes out next, you can actually plug in a memory system here.\nSo that has context.\nAnd the great thing about this is if we use for example Hermes agent OS like this, you can see that it actually saves the skill inside the Obsidian memory.\nWhy is that useful?\nIt's useful because for example, if we're switching between the memory system inside Hermes and then we go over to and we use, let's say, Deepseek.\nWell, it's got that memory stored from Hermes agent OS.\nThe same for example, if we go over to Claude Code over here and we select Quen 3.8 8 as the brain.\nWe can now use Quen 3.8 as the brain of Claude code but then also switch over to Hermes agent OS as well whenever we want to.\nAnd so you can describe what you want the 2.4 trillion parameter brain thinks it through.\nHermes creates whatever you ask it to do and then you can actually open that up in your browser.\nAlso some people say like this is difficult to set up a new model inside Hermes agent OS.\nYou can literally just go to the manage tab or your dashboard with Hermes agent.\nI would make sure that you update first.\nSo you can just click the update button and then from here you can create a new profile with Quen 3.8 Max plugged in.\nSo you can see here that we have a separate profile for Quen 3.8 ready to go.\nNow also when they were training Quen 3.8 Amax.\nThis model is very good at coding and actually ran 10 days autonomously with self-evolving development from an empty folder to production without handholding.\nSo basically it updated and you can trace what it's done inside the GitHub for this model and see exactly what it built and so it can code autonomously for hours.\nNow why is that important with Hermes agent?\nWell, essentially, if you're using the goal mode, and we have a separate section for goal mode over here, we can give it a goal and then it can go off and complete the goals and run autonomously for hours without us even having to touch it.\nSo, we can set the target and then just walk away with goal mode.\nAnd Quinn 3.8 Max is designed to actually do that.\nThe other benefit of this, and this is something I haven't tried yet, but would be super interesting, is you could use Hermes mixture of agents and you could use, for example, Fable 5 combined with Quen 3.8.\n8 max.\nAnd that fuses the answers together so you can get even better outputs than if you were coding with Quen 3.8 Max alone inside Hermes.\nAnd so if you combine the models, you could get even better outputs.\nNow, additionally, if you're coding with Quen 3.8, you can build some awesome stuff and you get Hermes agent to actually build that with Quen as a brain.\nSo, for example, we actually tested 50 different builds with this and it can create some amazing games, some apps, tools, websites, whatever you want directly with it.\nAnd so, whether you're using MCPS like I've shown you today, or for example, you're building out games with this, maybe you are trying to create a website, or maybe you're just using some of the tools like for example, forward/arn as we covered\nbefore, you can get very good outputs from this.\nHere's an actual promo video it created for us with Hermes agent, Quen 3.8 8 Max and then the skill of remotion.\nSo it can create animated videos.\nSo if we take a look at this, you can see it's super clean.\nIt looks really nice, very beautiful design.\nIt's actually very good at front end as well.\nAnd then this can create like a fully animated video for us with one single prompt.\nAnd so you can leave Quen 3.8 Max coding autonomously.\nYou give it a prompt to create a video like this and it can create something absolutely amazing.\nNow what we've actually got inside this agent OS is we have the video section here.\nSo we can type in a prompt then it can generate the B-roll, the script, the research, put it all together, edit it beautifully.\nAnd also the great thing about this is we have our workspace here where we can come back to it and see what we've built and watch it back later.\nAnd that's fantastic because what we can do is make sure that we never lose a project again.\nWe never lose something that we've coded and we can watch back and see what it's built which is super useful.\nSo that's basically how to use Quen 3.8 Max.\nIf you want to switch the model inside Hermes, you can actually go over to the manage tab here.\nAnd then from here, you would change the model.\nAnd you can see if you type in Quen 3.8 inside the set main model section, you can select it inside news portal and you can choose Quen 3.8 Max there or you can go over to open router and select Quen 3.8 Max there as well.\nSo super easy to switch the models.\nAnd also Hermes have just released version 0.2 which actually makes it faster and better with voice too.\nSo if you want to get our full agent operating system with Hermes, all the systems I've shown you today, you got Clawude working side by side here it the full memory system as well and everything else I've shown you.\nYou can get that inside the AI profit boardroom link in the comments description or just go to the AI profit bomb.com.\nThis is my AI automation community that shows you how to train agents set up systems like this and it gives you the full Hermes agent operating system as well.\nSo if you go to the classroom here then new daily updates.\nYou'll find the agent OS with a video tutorial.\nIt's last update date and a zip file to install it.\nWe also got a full 1hour course on Quen 3.8 Max if you want a full master class on it.\nAdditionally, we have loads of great trainings on Hermes agent as well.\nAnd you can jump on weekly coaching calls where you can share your screen, ask questions, get help in real time.\nInside the community, you can ask questions as well, and I personally answer them with a video tutorial every single day.\nYou can also direct message me whenever you need help.\nInside the map, you can meet people in your local area who are building out with agents like Hermes and Quen 3.8 Max.\nAnd that's all available inside the air profit boarding link in the comments description or just go to the aiprofit.com.",
  "language": "en",
  "duration": 468.694813,
  "segments": [
    {
      "start": 0.08,
      "end": 1.934,
      "text": "Quen 3.8 Max just dropped.",
      "source": "source_caption"
    },
    {
      "start": 1.934,
      "end": 6.309,
      "text": "I've already plugged it into Hermes Agent and it's pretty powerful from what I've tested so far.",
      "source": "source_caption"
    },
    {
      "start": 6.309,
      "end": 16.07,
      "text": "So, what I'm going to show you today is how to use Quen 3.8 Max with Hermes Agent and also how to get the most out of it and some of the cool stuff that we've created with it so far.",
      "source": "source_caption"
    },
    {
      "start": 16.08,
      "end": 29.002,
      "text": "So, the great thing about Quen 3.8 Max is that it's basically Fable 5 level and you can use your coding plan with it to plug it in directly into Hermes, which you actually can't do with Fable 5 unless you use the API.",
      "source": "source_caption"
    },
    {
      "start": 29.002,
      "end": 35.389,
      "text": "It's also a 2.4 trillion parameter brain and it actually beats Fable 5 on some of the key benchmarks.",
      "source": "source_caption"
    },
    {
      "start": 35.389,
      "end": 40.415,
      "text": "So basically Hermes agent is the hands and then you can change the brain.",
      "source": "source_caption"
    },
    {
      "start": 40.415,
      "end": 42.63,
      "text": "So you can plug in Quen 3.8 Max.",
      "source": "source_caption"
    },
    {
      "start": 42.64,
      "end": 50.275,
      "text": "This was a 2.4 trillion parameters, 1 million context window and also designed for agentic tasks.",
      "source": "source_caption"
    },
    {
      "start": 50.275,
      "end": 53.04,
      "text": "I'll show you some things we built with it so far.",
      "source": "source_caption"
    },
    {
      "start": 53.04,
      "end": 56.49,
      "text": "So, it created like some pretty cool coding apps as you can see right here.",
      "source": "source_caption"
    },
    {
      "start": 56.49,
      "end": 59.652,
      "text": "But also, we actually linked it to MCP.",
      "source": "source_caption"
    },
    {
      "start": 59.652,
      "end": 63.51,
      "text": "So, we linked it to Hicksfield MCP just to test it out and see what it could do.",
      "source": "source_caption"
    },
    {
      "start": 63.52,
      "end": 76.194,
      "text": "And we actually used and gave it our book image as an example to see if it could code and build some cool stuff with Higsfield MCP and Quen 3.8 Max inside Hermes agent.",
      "source": "source_caption"
    },
    {
      "start": 76.194,
      "end": 81.239,
      "text": "So, this is some of the stuff that he created, which is pretty nice and it was super easy to use.",
      "source": "source_caption"
    },
    {
      "start": 81.239,
      "end": 84.878,
      "text": "The other thing I would say about Quen 3.8 Max, it's pretty fast to respond.",
      "source": "source_caption"
    },
    {
      "start": 84.878,
      "end": 86.711,
      "text": "Here's another example.",
      "source": "source_caption"
    },
    {
      "start": 86.711,
      "end": 92.331,
      "text": "So, what we actually have inside our Hermes agentic operating system is all of our different profiles with different models plugged in.",
      "source": "source_caption"
    },
    {
      "start": 92.331,
      "end": 96.132,
      "text": "Now, one thing I'll say about Quen 3.0 Max is it's very fast to apply.",
      "source": "source_caption"
    },
    {
      "start": 96.132,
      "end": 100.233,
      "text": "So, we say, for example, learn and then give it access to a guide.",
      "source": "source_caption"
    },
    {
      "start": 100.233,
      "end": 102.802,
      "text": "And you can see here it says skill learned and registered.",
      "source": "source_caption"
    },
    {
      "start": 102.802,
      "end": 103.764,
      "text": "Here's a skill.",
      "source": "source_caption"
    },
    {
      "start": 103.764,
      "end": 108.207,
      "text": "Here's the path that saved locally as a skill MD file, which we can open up in a second.",
      "source": "source_caption"
    },
    {
      "start": 108.207,
      "end": 109.673,
      "text": "And here's what it captures.",
      "source": "source_caption"
    },
    {
      "start": 109.673,
      "end": 113.011,
      "text": "And so we've taught it a new skill in the space of a couple of minutes.",
      "source": "source_caption"
    },
    {
      "start": 113.011,
      "end": 118.997,
      "text": "And if we open this up, we've got the full skill that's written here based on the guide we actually trained it with.",
      "source": "source_caption"
    },
    {
      "start": 118.997,
      "end": 129.559,
      "text": "Now, one thing I definitely recommend with this is that when you're using new models, whether that's Fable 5 or whether that's Quinn 3.8 Max or whatever comes out next, you can actually plug in a memory system here.",
      "source": "source_caption"
    },
    {
      "start": 129.559,
      "end": 130.56,
      "text": "So that has context.",
      "source": "source_caption"
    },
    {
      "start": 130.56,
      "end": 139.395,
      "text": "And the great thing about this is if we use for example Hermes agent OS like this, you can see that it actually saves the skill inside the Obsidian memory.",
      "source": "source_caption"
    },
    {
      "start": 139.395,
      "end": 140.333,
      "text": "Why is that useful?",
      "source": "source_caption"
    },
    {
      "start": 140.333,
      "end": 149.911,
      "text": "It's useful because for example, if we're switching between the memory system inside Hermes and then we go over to and we use, let's say, Deepseek.",
      "source": "source_caption"
    },
    {
      "start": 149.911,
      "end": 152.828,
      "text": "Well, it's got that memory stored from Hermes agent OS.",
      "source": "source_caption"
    },
    {
      "start": 152.828,
      "end": 159.463,
      "text": "The same for example, if we go over to Claude Code over here and we select Quen 3.8 8 as the brain.",
      "source": "source_caption"
    },
    {
      "start": 159.463,
      "end": 167.396,
      "text": "We can now use Quen 3.8 as the brain of Claude code but then also switch over to Hermes agent OS as well whenever we want to.",
      "source": "source_caption"
    },
    {
      "start": 167.396,
      "end": 173.11,
      "text": "And so you can describe what you want the 2.4 trillion parameter brain thinks it through.",
      "source": "source_caption"
    },
    {
      "start": 173.12,
      "end": 178.067,
      "text": "Hermes creates whatever you ask it to do and then you can actually open that up in your browser.",
      "source": "source_caption"
    },
    {
      "start": 178.067,
      "end": 182.868,
      "text": "Also some people say like this is difficult to set up a new model inside Hermes agent OS.",
      "source": "source_caption"
    },
    {
      "start": 182.868,
      "end": 187.769,
      "text": "You can literally just go to the manage tab or your dashboard with Hermes agent.",
      "source": "source_caption"
    },
    {
      "start": 187.769,
      "end": 189.11,
      "text": "I would make sure that you update first.",
      "source": "source_caption"
    },
    {
      "start": 189.12,
      "end": 195.722,
      "text": "So you can just click the update button and then from here you can create a new profile with Quen 3.8 Max plugged in.",
      "source": "source_caption"
    },
    {
      "start": 195.722,
      "end": 201.558,
      "text": "So you can see here that we have a separate profile for Quen 3.8 ready to go.",
      "source": "source_caption"
    },
    {
      "start": 201.558,
      "end": 204.889,
      "text": "Now also when they were training Quen 3.8 Amax.",
      "source": "source_caption"
    },
    {
      "start": 204.889,
      "end": 216.52,
      "text": "This model is very good at coding and actually ran 10 days autonomously with self-evolving development from an empty folder to production without handholding.",
      "source": "source_caption"
    },
    {
      "start": 216.52,
      "end": 227.939,
      "text": "So basically it updated and you can trace what it's done inside the GitHub for this model and see exactly what it built and so it can code autonomously for hours.",
      "source": "source_caption"
    },
    {
      "start": 227.939,
      "end": 230.599,
      "text": "Now why is that important with Hermes agent?",
      "source": "source_caption"
    },
    {
      "start": 230.599,
      "end": 242.171,
      "text": "Well, essentially, if you're using the goal mode, and we have a separate section for goal mode over here, we can give it a goal and then it can go off and complete the goals and run autonomously for hours without us even having to touch it.",
      "source": "source_caption"
    },
    {
      "start": 242.171,
      "end": 245.871,
      "text": "So, we can set the target and then just walk away with goal mode.",
      "source": "source_caption"
    },
    {
      "start": 245.871,
      "end": 248.873,
      "text": "And Quinn 3.8 Max is designed to actually do that.",
      "source": "source_caption"
    },
    {
      "start": 248.873,
      "end": 259.56,
      "text": "The other benefit of this, and this is something I haven't tried yet, but would be super interesting, is you could use Hermes mixture of agents and you could use, for example, Fable 5 combined with Quen 3.8.",
      "source": "source_caption"
    },
    {
      "start": 259.56,
      "end": 260.068,
      "text": "8 max.",
      "source": "source_caption"
    },
    {
      "start": 260.068,
      "end": 271.191,
      "text": "And that fuses the answers together so you can get even better outputs than if you were coding with Quen 3.8 Max alone inside Hermes.",
      "source": "source_caption"
    },
    {
      "start": 271.191,
      "end": 275.262,
      "text": "And so if you combine the models, you could get even better outputs.",
      "source": "source_caption"
    },
    {
      "start": 275.262,
      "end": 284.452,
      "text": "Now, additionally, if you're coding with Quen 3.8, you can build some awesome stuff and you get Hermes agent to actually build that with Quen as a brain.",
      "source": "source_caption"
    },
    {
      "start": 284.452,
      "end": 292.497,
      "text": "So, for example, we actually tested 50 different builds with this and it can create some amazing games, some apps, tools, websites, whatever you want directly with it.",
      "source": "source_caption"
    },
    {
      "start": 292.497,
      "end": 305.03,
      "text": "And so, whether you're using MCPS like I've shown you today, or for example, you're building out games with this, maybe you are trying to create a website, or maybe you're just using some of the tools like for example, forward/arn as we covered",
      "source": "source_caption"
    },
    {
      "start": 305.04,
      "end": 308.583,
      "text": "before, you can get very good outputs from this.",
      "source": "source_caption"
    },
    {
      "start": 308.583,
      "end": 316.783,
      "text": "Here's an actual promo video it created for us with Hermes agent, Quen 3.8 8 Max and then the skill of remotion.",
      "source": "source_caption"
    },
    {
      "start": 316.783,
      "end": 318.405,
      "text": "So it can create animated videos.",
      "source": "source_caption"
    },
    {
      "start": 318.405,
      "end": 321.382,
      "text": "So if we take a look at this, you can see it's super clean.",
      "source": "source_caption"
    },
    {
      "start": 321.382,
      "end": 323.626,
      "text": "It looks really nice, very beautiful design.",
      "source": "source_caption"
    },
    {
      "start": 323.626,
      "end": 325.59,
      "text": "It's actually very good at front end as well.",
      "source": "source_caption"
    },
    {
      "start": 325.6,
      "end": 331.067,
      "text": "And then this can create like a fully animated video for us with one single prompt.",
      "source": "source_caption"
    },
    {
      "start": 331.067,
      "end": 335.494,
      "text": "And so you can leave Quen 3.8 Max coding autonomously.",
      "source": "source_caption"
    },
    {
      "start": 335.494,
      "end": 340.374,
      "text": "You give it a prompt to create a video like this and it can create something absolutely amazing.",
      "source": "source_caption"
    },
    {
      "start": 340.374,
      "end": 344.794,
      "text": "Now what we've actually got inside this agent OS is we have the video section here.",
      "source": "source_caption"
    },
    {
      "start": 344.794,
      "end": 351.857,
      "text": "So we can type in a prompt then it can generate the B-roll, the script, the research, put it all together, edit it beautifully.",
      "source": "source_caption"
    },
    {
      "start": 351.857,
      "end": 358.74,
      "text": "And also the great thing about this is we have our workspace here where we can come back to it and see what we've built and watch it back later.",
      "source": "source_caption"
    },
    {
      "start": 358.74,
      "end": 363.99,
      "text": "And that's fantastic because what we can do is make sure that we never lose a project again.",
      "source": "source_caption"
    },
    {
      "start": 364.0,
      "end": 368.568,
      "text": "We never lose something that we've coded and we can watch back and see what it's built which is super useful.",
      "source": "source_caption"
    },
    {
      "start": 368.568,
      "end": 372.283,
      "text": "So that's basically how to use Quen 3.8 Max.",
      "source": "source_caption"
    },
    {
      "start": 372.283,
      "end": 376.915,
      "text": "If you want to switch the model inside Hermes, you can actually go over to the manage tab here.",
      "source": "source_caption"
    },
    {
      "start": 376.915,
      "end": 379.832,
      "text": "And then from here, you would change the model.",
      "source": "source_caption"
    },
    {
      "start": 379.832,
      "end": 394.422,
      "text": "And you can see if you type in Quen 3.8 inside the set main model section, you can select it inside news portal and you can choose Quen 3.8 Max there or you can go over to open router and select Quen 3.8 Max there as well.",
      "source": "source_caption"
    },
    {
      "start": 394.422,
      "end": 396.128,
      "text": "So super easy to switch the models.",
      "source": "source_caption"
    },
    {
      "start": 396.128,
      "end": 402.497,
      "text": "And also Hermes have just released version 0.2 which actually makes it faster and better with voice too.",
      "source": "source_caption"
    },
    {
      "start": 402.497,
      "end": 411.996,
      "text": "So if you want to get our full agent operating system with Hermes, all the systems I've shown you today, you got Clawude working side by side here it the full memory system as well and everything else I've shown you.",
      "source": "source_caption"
    },
    {
      "start": 411.996,
      "end": 417.996,
      "text": "You can get that inside the AI profit boardroom link in the comments description or just go to the AI profit bomb.com.",
      "source": "source_caption"
    },
    {
      "start": 417.996,
      "end": 426.088,
      "text": "This is my AI automation community that shows you how to train agents set up systems like this and it gives you the full Hermes agent operating system as well.",
      "source": "source_caption"
    },
    {
      "start": 426.088,
      "end": 429.141,
      "text": "So if you go to the classroom here then new daily updates.",
      "source": "source_caption"
    },
    {
      "start": 429.141,
      "end": 431.805,
      "text": "You'll find the agent OS with a video tutorial.",
      "source": "source_caption"
    },
    {
      "start": 431.805,
      "end": 434.297,
      "text": "It's last update date and a zip file to install it.",
      "source": "source_caption"
    },
    {
      "start": 434.297,
      "end": 439.581,
      "text": "We also got a full 1hour course on Quen 3.8 Max if you want a full master class on it.",
      "source": "source_caption"
    },
    {
      "start": 439.581,
      "end": 443.549,
      "text": "Additionally, we have loads of great trainings on Hermes agent as well.",
      "source": "source_caption"
    },
    {
      "start": 443.549,
      "end": 448.723,
      "text": "And you can jump on weekly coaching calls where you can share your screen, ask questions, get help in real time.",
      "source": "source_caption"
    },
    {
      "start": 448.723,
      "end": 453.848,
      "text": "Inside the community, you can ask questions as well, and I personally answer them with a video tutorial every single day.",
      "source": "source_caption"
    },
    {
      "start": 453.848,
      "end": 456.081,
      "text": "You can also direct message me whenever you need help.",
      "source": "source_caption"
    },
    {
      "start": 456.081,
      "end": 461.794,
      "text": "Inside the map, you can meet people in your local area who are building out with agents like Hermes and Quen 3.8 Max.",
      "source": "source_caption"
    },
    {
      "start": 461.794,
      "end": 469.8,
      "text": "And that's all available inside the air profit boarding link in the comments description or just go to the aiprofit.com.",
      "source": "source_caption"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "en",
    "duration": 468.694813,
    "reference_path": "/Users/sagawa/AI_WORK/video_notes/url/20260805_234558__Hermes_Agent_Qwen_3.8_Max_is_INSANE/reference_subtitle.vtt",
    "reference_kind": "automatic_original",
    "reference_language": "en-orig",
    "selected_source": "youtube_auto_original",
    "decision_reason": "trusted_full_source_selected_after_cross_check",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 87,
      "first_start": 0.08,
      "final_end": 469.8,
      "duration_coverage": 1.0,
      "maximum_gap_seconds": 0.08,
      "inverted_count": 0,
      "language": {
        "requested": "en",
        "matched": true,
        "latin": 6430,
        "han": 0,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "whisper_profile": {
      "segment_count": 86,
      "first_start": 0.0,
      "final_end": 467.96,
      "duration_coverage": 0.998432,
      "maximum_gap_seconds": 0.68,
      "inverted_count": 0,
      "language": {
        "requested": "en",
        "matched": true,
        "latin": 6451,
        "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": 16,
      "median_score": 0.9282,
      "divergent_window_count": 0,
      "divergent_ratio": 0.0,
      "divergent_windows": []
    }
  }
}