{
  "text": "大家好,Deepseek V4 Flash 正式版发布之后,得到了非常大的关注。很多人用不同的AI工具去对接Deepseek V4 Flash。\n那么也有很多讨论,这个工具可能调用Deepseek更省钱,或者效率更高。\n本期视频就来测试一下,在不同的AI变成工具里面同样的任务,使用Deepseek V4 Flash 正式版所消耗的时间和成本分别是多少。\n最后我们再总结一下,哪一个AI变成工具比较适合来使用Deepseek V4 Flash。\n那么首先在Deepseek官网推荐的AI变成工具,或者是A型的工具,有以下几个,比如Cloud Code, Codex, OpenCode,还有几个个人助手类的。\n然后我这边测试的话,使用的是4个AI变成工具,第一个是Cloud Code,第二个是Codex App,还有OpenCode以及Pi。\n那么这4个基本上涵盖了目前大家使用最多的这几款工具。\n还有一个工具叫OMP的,那么是基于Pi这个框架来进行迭代开发了。\n所以我就不再测试了。\n那么测试提示词,是用工具来制作一款这样的CS1.6的这种游戏。\n这个游戏是一个3D的游戏,提示词也是都一样的。\n每一个AI变成工具,接受到提示词都是一样的。\n测试流程就是,我希望使用同样提示词,达到一个差不多的效果。\n然后来算一下它花费的时间和成本。\n那这个过程是包含了从0到1去生成整个项目。\n然后来对这个项目进行一些微调,因为可能会有一些问题。\n然后最后达到一个可玩的一个这样的效果。\n最终来计算一下花费的时间,花费的成本,\n以及最终的结果的一个评分。\n看一下谁是最好的。\n我们来看一下最终的效果是什么样子。\n我们先看一下Cloud Code。\n那这边的话我们打开一下我们的测试的命令。\n我们可以看一下整个的一个效果。\n那么它是一个这样的,我们就随便点一下。\n主要是看一下这个效果。\n那其实这几个AI变成工具生成效果都非常一般。\n都没有之前的其他模型生成效果更好。\n所以整个来说,射击什么都是正常。\nOK,我们看第二个。\n我们来看第二个。\n第二个是Open Code来实现的一个效果。\n我们来启动一下。\n然后这边的话是页面效果也还行。\n但是这个枪支的效果好像很一般。\n然后的话我们买一个武器。\n这个射击的效果就非常一般了。\nOK,我们也是到这里就结束。\n我们再看一下使用Pi生成的效果进入。\n然后这边的话我们随便买一个。\nOK,它这个顺序还反了。\n它这个没法射击还是有些问题的。\n我们再看一下Cutis App做出来的效果。\n这个地图很小。\n然后的话我们可以看一下。\nOK,我们看一下是不是这个效果。\n这个效果也真很一般。\n那么从上面的测试结果来看。\n虽然我们没有测试完整个流程。\n那么从页面效果交互。\n这一类的这个对比来看。\nCloud Code的效果是最好的。\n那其次是Open Code。\n然后再就是Cutis App和Pi。\n那么在使用成本上。\n我给每一个AI变成工具都创建了一个独有的API Key。\n从这个DeepSick后台的这个成本分析里面。\n我们可以看到啊。\nPi是使用的成本是最高的。\n1.7亿元。\n其次是这个CodeS App。\n接下来就是Cloud Code和Open Code。\n那么这个结果非常出乎我的意料。\n那么我认为Pi这个框架。\n加上DeepSick Flash应该属于又快又省钱。\n但是呢从测试结果来看。\nPi使用了60分钟。\n然后成本是最高的。\n时间是最长的。\n而且效果是最差的。\n目前还不知道具体的原因是什么。\n然后CodeS App也是使用了比较长的时间。\n而且效果也非常差。\n可能跟它的内置的验证流程是否有关系。\n因为我们其实使用的是第三方的模型来对接Class App。\n相当于不属于官方的一个范围。\n然后表现最好的是Cloud Code。\n这个应该是在我的意料中的。\n毕竟它有那么一套成熟的一个流程。\n那么我在测试过程中也是把所有的技能都全部禁用掉了。\n然后基本上用纯内置提示词加上这个模型的能力来跑这样一个结果。\n所以从我这个测试结果啊。\n虽然说不代表一个权威啊。\n来看还是推荐大家使用Cloud Code去对接DeepSick Flash。\n会比较好一点。\n毕竟生态比较丰富嘛。\n然后或者说你使用Cloud X也行。\n那么你可以推荐你使用这个Open Code。\n那么其实不管是Cloud Code还是Cloud App。\n他们的内置的提示词或者流程,\n肯定是优先适配于自己的模型。\n那比如说在之前的视频里面我分享过了。\nCloud Code在OPUS5.0之后会减省80%的内置提示词。\n那么针对非OPUS5.0的模型。\n他们会用其他的这种提示词。\n那么可能当DeepSick的能力超过OPUS5.0。\n那么再使用Cloud Code的时候。\n那么就会用老的提示词用到乔模型上。\n那效果可能会打折。\n所以最好的还是模型公司自己出了这个AI变成工具。\n或者我们叫Hannis工具。\nDepSick官方最近也在内测自己的这个Hannis工具。\n也非常期待它的正式上线。\n到时候我们再来测试一下。\n然后这几个工具的话。\n如果你喜欢自己折腾。\n那希望自己去配置单独的包。\n或者说第三方的扩展。\n那么我非常推荐使用Pi。\n因为Pi其实是一个非常简单的一个Agent工具。\n它不仅可以开发自己的Agent工具。\n使用它来进行编程。\n当然它的官网也有非常的第三方包。\n比如说像子代理,像动态工作流,像目标。\n以及其他的非常好用的一些包。\n那你把它按需要去加载进来。\n那也是一个非常好用的工具。\n那么视频就到这儿。\n希望这个视频对你有所帮助。",
  "language": "zh",
  "duration": 339.081,
  "segments": [
    {
      "id": 0,
      "seek": 0,
      "start": 0.0,
      "end": 8.700000000000001,
      "text": "大家好,Deepseek V4 Flash 正式版发布之后,得到了非常大的关注。很多人用不同的AI工具去对接Deepseek V4 Flash。",
      "tokens": [
        50365,
        15248,
        11,
        11089,
        595,
        405,
        916,
        691,
        19,
        20232,
        220,
        15789,
        27584,
        42096,
        28926,
        34688,
        9574,
        13547,
        11,
        5916,
        21381,
        14392,
        39156,
        28053,
        26432,
        1543,
        20778,
        4035,
        9254,
        47123,
        1546,
        48698,
        23323,
        39806,
        6734,
        8713,
        14468,
        11089,
        595,
        405,
        916,
        691,
        19,
        20232,
        1543,
        50800
      ],
      "temperature": 0,
      "avg_logprob": -0.18727501460484097,
      "compression_ratio": 1.2895622895622896,
      "no_speech_prob": 9.500745676294109e-12,
      "source": "whisper"
    },
    {
      "id": 1,
      "seek": 0,
      "start": 8.700000000000001,
      "end": 14.08,
      "text": "那么也有很多讨论,这个工具可能调用Deepseek更省钱,或者效率更高。",
      "tokens": [
        50800,
        35693,
        6404,
        2412,
        20778,
        7422,
        101,
        7422,
        118,
        11,
        15368,
        23323,
        39806,
        16657,
        8897,
        225,
        9254,
        11089,
        595,
        405,
        916,
        19002,
        2862,
        223,
        39623,
        11,
        31148,
        43076,
        44866,
        19002,
        12979,
        1543,
        51069
      ],
      "temperature": 0,
      "avg_logprob": -0.18727501460484097,
      "compression_ratio": 1.2895622895622896,
      "no_speech_prob": 9.500745676294109e-12,
      "source": "whisper"
    },
    {
      "id": 2,
      "seek": 0,
      "start": 14.08,
      "end": 24.400000000000002,
      "text": "本期视频就来测试一下,在不同的AI变成工具里面同样的任务,使用Deepseek V4 Flash 正式版所消耗的时间和成本分别是多少。",
      "tokens": [
        51069,
        8802,
        16786,
        40656,
        39752,
        3111,
        6912,
        11038,
        233,
        5233,
        243,
        8861,
        11,
        3581,
        47123,
        1546,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        15759,
        8833,
        13089,
        14496,
        1546,
        26443,
        5087,
        94,
        11,
        22982,
        9254,
        11089,
        595,
        405,
        916,
        691,
        19,
        20232,
        220,
        15789,
        27584,
        42096,
        5966,
        28837,
        4450,
        245,
        1546,
        44848,
        12565,
        11336,
        8802,
        6627,
        18453,
        1541,
        41492,
        1543,
        51585
      ],
      "temperature": 0,
      "avg_logprob": -0.18727501460484097,
      "compression_ratio": 1.2895622895622896,
      "no_speech_prob": 9.500745676294109e-12,
      "source": "whisper"
    },
    {
      "id": 3,
      "seek": 2440,
      "start": 24.4,
      "end": 29.4,
      "text": "最后我们再总结一下,哪一个AI变成工具比较适合来使用Deepseek V4 Flash。",
      "tokens": [
        50365,
        8661,
        13547,
        15003,
        8623,
        33440,
        45641,
        8861,
        11,
        17028,
        20182,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        11706,
        9830,
        225,
        2215,
        224,
        14245,
        6912,
        22982,
        9254,
        11089,
        595,
        405,
        916,
        691,
        19,
        20232,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.15956734774405495,
      "compression_ratio": 1.3734939759036144,
      "no_speech_prob": 1.1220783850685034e-11,
      "source": "whisper"
    },
    {
      "id": 4,
      "seek": 2440,
      "start": 29.4,
      "end": 40.4,
      "text": "那么首先在Deepseek官网推荐的AI变成工具,或者是A型的工具,有以下几个,比如Cloud Code, Codex, OpenCode,还有几个个人助手类的。",
      "tokens": [
        50615,
        35693,
        36490,
        3581,
        11089,
        595,
        405,
        916,
        31929,
        16469,
        239,
        33597,
        31409,
        238,
        1546,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        11,
        31148,
        1541,
        32,
        39823,
        1546,
        23323,
        39806,
        11,
        2412,
        3588,
        4438,
        6336,
        254,
        7549,
        11,
        36757,
        32787,
        15549,
        11,
        15549,
        87,
        11,
        7238,
        34,
        1429,
        11,
        35091,
        6336,
        254,
        7549,
        7549,
        4035,
        37618,
        11389,
        22113,
        119,
        1546,
        1543,
        51165
      ],
      "temperature": 0,
      "avg_logprob": -0.15956734774405495,
      "compression_ratio": 1.3734939759036144,
      "no_speech_prob": 1.1220783850685034e-11,
      "source": "whisper"
    },
    {
      "id": 5,
      "seek": 2440,
      "start": 40.4,
      "end": 48.4,
      "text": "然后我这边测试的话,使用的是4个AI变成工具,第一个是Cloud Code,第二个是Codex App,还有OpenCode以及Pi。",
      "tokens": [
        51165,
        26636,
        1654,
        5562,
        40503,
        11038,
        233,
        5233,
        243,
        44575,
        11,
        22982,
        9254,
        24620,
        19,
        7549,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        11,
        18049,
        7549,
        1541,
        32787,
        15549,
        11,
        19693,
        7549,
        1541,
        34,
        1429,
        87,
        3132,
        11,
        35091,
        45569,
        34,
        1429,
        40282,
        47,
        72,
        1543,
        51565
      ],
      "temperature": 0,
      "avg_logprob": -0.15956734774405495,
      "compression_ratio": 1.3734939759036144,
      "no_speech_prob": 1.1220783850685034e-11,
      "source": "whisper"
    },
    {
      "id": 6,
      "seek": 2440,
      "start": 48.4,
      "end": 53.4,
      "text": "那么这4个基本上涵盖了目前大家使用最多的这几款工具。",
      "tokens": [
        51565,
        35693,
        5562,
        19,
        7549,
        37946,
        5708,
        35681,
        113,
        5419,
        244,
        2289,
        39004,
        6868,
        22982,
        9254,
        8661,
        6392,
        1546,
        5562,
        6336,
        254,
        48798,
        23323,
        39806,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.15956734774405495,
      "compression_ratio": 1.3734939759036144,
      "no_speech_prob": 1.1220783850685034e-11,
      "source": "whisper"
    },
    {
      "id": 7,
      "seek": 5340,
      "start": 53.4,
      "end": 58.4,
      "text": "还有一个工具叫OMP的,那么是基于Pi这个框架来进行迭代开发了。",
      "tokens": [
        50365,
        35091,
        20182,
        23323,
        39806,
        19855,
        5251,
        47,
        1546,
        11,
        35693,
        1541,
        26008,
        37732,
        47,
        72,
        15368,
        21416,
        228,
        7360,
        114,
        6912,
        36700,
        8082,
        3316,
        255,
        19105,
        18937,
        28926,
        2289,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 8,
      "seek": 5340,
      "start": 58.4,
      "end": 60.4,
      "text": "所以我就不再测试了。",
      "tokens": [
        50615,
        7239,
        22020,
        1960,
        8623,
        11038,
        233,
        5233,
        243,
        2289,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 9,
      "seek": 5340,
      "start": 60.4,
      "end": 66.4,
      "text": "那么测试提示词,是用工具来制作一款这样的CS1.6的这种游戏。",
      "tokens": [
        50715,
        35693,
        11038,
        233,
        5233,
        243,
        20949,
        25696,
        5233,
        235,
        11,
        1541,
        9254,
        23323,
        39806,
        6912,
        25491,
        11914,
        2257,
        48798,
        21209,
        1546,
        26283,
        16,
        13,
        21,
        1546,
        5562,
        39810,
        9592,
        116,
        1486,
        237,
        1543,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 10,
      "seek": 5340,
      "start": 66.4,
      "end": 70.4,
      "text": "这个游戏是一个3D的游戏,提示词也是都一样的。",
      "tokens": [
        51015,
        15368,
        9592,
        116,
        1486,
        237,
        1541,
        20182,
        18,
        35,
        1546,
        9592,
        116,
        1486,
        237,
        11,
        20949,
        25696,
        5233,
        235,
        22021,
        7182,
        2257,
        14496,
        1546,
        1543,
        51215
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 11,
      "seek": 5340,
      "start": 70.4,
      "end": 73.4,
      "text": "每一个AI变成工具,接受到提示词都是一样的。",
      "tokens": [
        51215,
        23664,
        20182,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        11,
        14468,
        23151,
        4511,
        20949,
        25696,
        5233,
        235,
        22796,
        2257,
        14496,
        1546,
        1543,
        51365
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 12,
      "seek": 5340,
      "start": 73.4,
      "end": 78.4,
      "text": "测试流程就是,我希望使用同样提示词,达到一个差不多的效果。",
      "tokens": [
        51365,
        11038,
        233,
        5233,
        243,
        27854,
        29649,
        5620,
        11,
        1654,
        29955,
        22982,
        9254,
        13089,
        14496,
        20949,
        25696,
        5233,
        235,
        11,
        9830,
        122,
        4511,
        20182,
        37876,
        1546,
        43076,
        9319,
        1543,
        51615
      ],
      "temperature": 0,
      "avg_logprob": -0.13066338011937112,
      "compression_ratio": 1.407017543859649,
      "no_speech_prob": 1.65525215062301e-11,
      "source": "whisper"
    },
    {
      "id": 13,
      "seek": 7840,
      "start": 78.4,
      "end": 81.4,
      "text": "然后来算一下它花费的时间和成本。",
      "tokens": [
        50365,
        26636,
        6912,
        19497,
        8861,
        11284,
        20127,
        18464,
        117,
        1546,
        44848,
        12565,
        11336,
        8802,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 14,
      "seek": 7840,
      "start": 81.4,
      "end": 84.4,
      "text": "那这个过程是包含了从0到1去生成整个项目。",
      "tokens": [
        50515,
        4184,
        15368,
        16866,
        29649,
        1541,
        23305,
        2392,
        104,
        2289,
        35630,
        15,
        4511,
        16,
        6734,
        8244,
        11336,
        27662,
        7549,
        10178,
        117,
        11386,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 15,
      "seek": 7840,
      "start": 84.4,
      "end": 88.4,
      "text": "然后来对这个项目进行一些微调,因为可能会有一些问题。",
      "tokens": [
        50665,
        26636,
        6912,
        8713,
        15368,
        10178,
        117,
        11386,
        36700,
        8082,
        38515,
        39152,
        8897,
        225,
        11,
        34627,
        16657,
        12949,
        32241,
        13824,
        34069,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 16,
      "seek": 7840,
      "start": 88.4,
      "end": 91.4,
      "text": "然后最后达到一个可玩的一个这样的效果。",
      "tokens": [
        50865,
        26636,
        8661,
        13547,
        9830,
        122,
        4511,
        20182,
        4429,
        19912,
        1546,
        20182,
        21209,
        1546,
        43076,
        9319,
        1543,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 17,
      "seek": 7840,
      "start": 91.4,
      "end": 95.4,
      "text": "最终来计算一下花费的时间,花费的成本,",
      "tokens": [
        51015,
        8661,
        10115,
        230,
        6912,
        7422,
        94,
        19497,
        8861,
        20127,
        18464,
        117,
        1546,
        44848,
        11,
        20127,
        18464,
        117,
        1546,
        11336,
        8802,
        11,
        51215
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 18,
      "seek": 7840,
      "start": 95.4,
      "end": 97.4,
      "text": "以及最终的结果的一个评分。",
      "tokens": [
        51215,
        40282,
        8661,
        10115,
        230,
        1546,
        45641,
        9319,
        1546,
        20182,
        5233,
        226,
        6627,
        1543,
        51315
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 19,
      "seek": 7840,
      "start": 97.4,
      "end": 99.4,
      "text": "看一下谁是最好的。",
      "tokens": [
        51315,
        28324,
        33556,
        1541,
        8661,
        20715,
        1543,
        51415
      ],
      "temperature": 0,
      "avg_logprob": -0.09103097102438756,
      "compression_ratio": 1.4534412955465588,
      "no_speech_prob": 1.7616536701625662e-11,
      "source": "whisper"
    },
    {
      "id": 20,
      "seek": 9940,
      "start": 99.4,
      "end": 102.4,
      "text": "我们来看一下最终的效果是什么样子。",
      "tokens": [
        50365,
        15003,
        6912,
        28324,
        8661,
        10115,
        230,
        1546,
        43076,
        9319,
        1541,
        10440,
        14496,
        7626,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 21,
      "seek": 9940,
      "start": 102.4,
      "end": 103.4,
      "text": "我们先看一下Cloud Code。",
      "tokens": [
        50515,
        15003,
        10108,
        28324,
        32787,
        15549,
        1543,
        50565
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 22,
      "seek": 9940,
      "start": 103.4,
      "end": 106.4,
      "text": "那这边的话我们打开一下我们的测试的命令。",
      "tokens": [
        50565,
        4184,
        5562,
        40503,
        44575,
        15003,
        12467,
        18937,
        8861,
        15003,
        1546,
        11038,
        233,
        5233,
        243,
        1546,
        25236,
        49061,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 23,
      "seek": 9940,
      "start": 106.4,
      "end": 108.4,
      "text": "我们可以看一下整个的一个效果。",
      "tokens": [
        50715,
        15003,
        6723,
        28324,
        27662,
        7549,
        1546,
        20182,
        43076,
        9319,
        1543,
        50815
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 24,
      "seek": 9940,
      "start": 108.4,
      "end": 111.4,
      "text": "那么它是一个这样的,我们就随便点一下。",
      "tokens": [
        50815,
        35693,
        11284,
        1541,
        20182,
        21209,
        1546,
        11,
        15003,
        3111,
        10673,
        237,
        27364,
        12579,
        8861,
        1543,
        50965
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 25,
      "seek": 9940,
      "start": 111.4,
      "end": 113.4,
      "text": "主要是看一下这个效果。",
      "tokens": [
        50965,
        13557,
        4275,
        1541,
        28324,
        15368,
        43076,
        9319,
        1543,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 26,
      "seek": 9940,
      "start": 113.4,
      "end": 116.4,
      "text": "那其实这几个AI变成工具生成效果都非常一般。",
      "tokens": [
        51065,
        4184,
        41646,
        5562,
        6336,
        254,
        7549,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        8244,
        11336,
        43076,
        9319,
        7182,
        14392,
        2257,
        49640,
        1543,
        51215
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 27,
      "seek": 9940,
      "start": 116.4,
      "end": 120.4,
      "text": "都没有之前的其他模型生成效果更好。",
      "tokens": [
        51215,
        7182,
        17944,
        32442,
        1546,
        9572,
        5000,
        41908,
        39823,
        8244,
        11336,
        43076,
        9319,
        19002,
        2131,
        1543,
        51415
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 28,
      "seek": 9940,
      "start": 120.4,
      "end": 122.4,
      "text": "所以整个来说,射击什么都是正常。",
      "tokens": [
        51415,
        7239,
        27662,
        7549,
        6912,
        8090,
        11,
        1530,
        226,
        6336,
        119,
        10440,
        22796,
        15789,
        11279,
        1543,
        51515
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 29,
      "seek": 9940,
      "start": 122.4,
      "end": 123.4,
      "text": "OK,我们看第二个。",
      "tokens": [
        51515,
        9443,
        11,
        15003,
        4200,
        19693,
        7549,
        1543,
        51565
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 30,
      "seek": 9940,
      "start": 123.4,
      "end": 124.4,
      "text": "我们来看第二个。",
      "tokens": [
        51565,
        15003,
        6912,
        4200,
        19693,
        7549,
        1543,
        51615
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 31,
      "seek": 9940,
      "start": 124.4,
      "end": 126.4,
      "text": "第二个是Open Code来实现的一个效果。",
      "tokens": [
        51615,
        19693,
        7549,
        1541,
        45569,
        15549,
        6912,
        24726,
        20204,
        1546,
        20182,
        43076,
        9319,
        1543,
        51715
      ],
      "temperature": 0,
      "avg_logprob": -0.10717724109518117,
      "compression_ratio": 1.6012084592145015,
      "no_speech_prob": 1.7191860782195256e-11,
      "source": "whisper"
    },
    {
      "id": 32,
      "seek": 12640,
      "start": 126.4,
      "end": 128.4,
      "text": "我们来启动一下。",
      "tokens": [
        50365,
        15003,
        6912,
        2392,
        107,
        34961,
        8861,
        1543,
        50465
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 33,
      "seek": 12640,
      "start": 128.4,
      "end": 131.4,
      "text": "然后这边的话是页面效果也还行。",
      "tokens": [
        50465,
        26636,
        5562,
        40503,
        44575,
        1541,
        10178,
        113,
        8833,
        43076,
        9319,
        6404,
        14852,
        8082,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 34,
      "seek": 12640,
      "start": 131.4,
      "end": 133.4,
      "text": "但是这个枪支的效果好像很一般。",
      "tokens": [
        50615,
        11189,
        15368,
        7360,
        103,
        24400,
        1546,
        43076,
        9319,
        33242,
        4563,
        2257,
        49640,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 35,
      "seek": 12640,
      "start": 133.4,
      "end": 136.4,
      "text": "然后的话我们买一个武器。",
      "tokens": [
        50715,
        26636,
        44575,
        15003,
        2930,
        108,
        20182,
        38194,
        34386,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 36,
      "seek": 12640,
      "start": 136.4,
      "end": 139.4,
      "text": "这个射击的效果就非常一般了。",
      "tokens": [
        50865,
        15368,
        1530,
        226,
        6336,
        119,
        1546,
        43076,
        9319,
        3111,
        14392,
        2257,
        49640,
        2289,
        1543,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 37,
      "seek": 12640,
      "start": 139.4,
      "end": 141.4,
      "text": "OK,我们也是到这里就结束。",
      "tokens": [
        51015,
        9443,
        11,
        15003,
        22021,
        4511,
        35102,
        3111,
        45641,
        44320,
        1543,
        51115
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 38,
      "seek": 12640,
      "start": 141.4,
      "end": 144.4,
      "text": "我们再看一下使用Pi生成的效果进入。",
      "tokens": [
        51115,
        15003,
        8623,
        28324,
        22982,
        9254,
        47,
        72,
        8244,
        11336,
        1546,
        43076,
        9319,
        36700,
        14028,
        1543,
        51265
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 39,
      "seek": 12640,
      "start": 144.4,
      "end": 147.4,
      "text": "然后这边的话我们随便买一个。",
      "tokens": [
        51265,
        26636,
        5562,
        40503,
        44575,
        15003,
        10673,
        237,
        27364,
        2930,
        108,
        20182,
        1543,
        51415
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 40,
      "seek": 12640,
      "start": 147.4,
      "end": 150.4,
      "text": "OK,它这个顺序还反了。",
      "tokens": [
        51415,
        9443,
        11,
        11284,
        15368,
        10178,
        118,
        6346,
        237,
        14852,
        22138,
        2289,
        1543,
        51565
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 41,
      "seek": 12640,
      "start": 150.4,
      "end": 155.4,
      "text": "它这个没法射击还是有些问题的。",
      "tokens": [
        51565,
        11284,
        15368,
        10062,
        11148,
        1530,
        226,
        6336,
        119,
        45726,
        2412,
        13824,
        34069,
        1546,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.10957661480970786,
      "compression_ratio": 1.5310077519379846,
      "no_speech_prob": 1.9892700334200875e-11,
      "source": "whisper"
    },
    {
      "id": 42,
      "seek": 15540,
      "start": 155.4,
      "end": 158.4,
      "text": "我们再看一下Cutis App做出来的效果。",
      "tokens": [
        50365,
        15003,
        8623,
        28324,
        34,
        325,
        271,
        3132,
        10907,
        44561,
        1546,
        43076,
        9319,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 43,
      "seek": 15540,
      "start": 158.4,
      "end": 160.4,
      "text": "这个地图很小。",
      "tokens": [
        50515,
        15368,
        10928,
        3919,
        122,
        4563,
        7322,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 44,
      "seek": 15540,
      "start": 160.4,
      "end": 163.4,
      "text": "然后的话我们可以看一下。",
      "tokens": [
        50615,
        26636,
        44575,
        15003,
        6723,
        28324,
        1543,
        50765
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 45,
      "seek": 15540,
      "start": 163.4,
      "end": 169.4,
      "text": "OK,我们看一下是不是这个效果。",
      "tokens": [
        50765,
        9443,
        11,
        15003,
        28324,
        23034,
        15368,
        43076,
        9319,
        1543,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 46,
      "seek": 15540,
      "start": 169.4,
      "end": 172.4,
      "text": "这个效果也真很一般。",
      "tokens": [
        51065,
        15368,
        43076,
        9319,
        6404,
        6303,
        4563,
        2257,
        49640,
        1543,
        51215
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 47,
      "seek": 15540,
      "start": 172.4,
      "end": 177.4,
      "text": "那么从上面的测试结果来看。",
      "tokens": [
        51215,
        35693,
        35630,
        49750,
        1546,
        11038,
        233,
        5233,
        243,
        45641,
        9319,
        6912,
        4200,
        1543,
        51465
      ],
      "temperature": 0,
      "avg_logprob": -0.3352887596882565,
      "compression_ratio": 1.2342857142857142,
      "no_speech_prob": 1.8228108258910858e-11,
      "source": "whisper"
    },
    {
      "id": 48,
      "seek": 17740,
      "start": 177.4,
      "end": 179.4,
      "text": "虽然我们没有测试完整个流程。",
      "tokens": [
        50365,
        12026,
        121,
        5823,
        15003,
        17944,
        11038,
        233,
        5233,
        243,
        14128,
        27662,
        7549,
        27854,
        29649,
        1543,
        50465
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 49,
      "seek": 17740,
      "start": 179.4,
      "end": 181.4,
      "text": "那么从页面效果交互。",
      "tokens": [
        50465,
        35693,
        35630,
        10178,
        113,
        8833,
        43076,
        9319,
        28455,
        1369,
        240,
        1543,
        50565
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 50,
      "seek": 17740,
      "start": 181.4,
      "end": 183.4,
      "text": "这一类的这个对比来看。",
      "tokens": [
        50565,
        5562,
        2257,
        22113,
        119,
        1546,
        15368,
        8713,
        11706,
        6912,
        4200,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 51,
      "seek": 17740,
      "start": 183.4,
      "end": 185.4,
      "text": "Cloud Code的效果是最好的。",
      "tokens": [
        50665,
        32787,
        15549,
        1546,
        43076,
        9319,
        1541,
        8661,
        20715,
        1543,
        50765
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 52,
      "seek": 17740,
      "start": 185.4,
      "end": 186.4,
      "text": "那其次是Open Code。",
      "tokens": [
        50765,
        4184,
        9572,
        9487,
        1541,
        45569,
        15549,
        1543,
        50815
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 53,
      "seek": 17740,
      "start": 186.4,
      "end": 188.4,
      "text": "然后再就是Cutis App和Pi。",
      "tokens": [
        50815,
        26636,
        8623,
        5620,
        34,
        325,
        271,
        3132,
        12565,
        47,
        72,
        1543,
        50915
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 54,
      "seek": 17740,
      "start": 188.4,
      "end": 190.4,
      "text": "那么在使用成本上。",
      "tokens": [
        50915,
        35693,
        3581,
        22982,
        9254,
        11336,
        8802,
        5708,
        1543,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 55,
      "seek": 17740,
      "start": 190.4,
      "end": 194.4,
      "text": "我给每一个AI变成工具都创建了一个独有的API Key。",
      "tokens": [
        51015,
        1654,
        23197,
        23664,
        20182,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        7182,
        2437,
        249,
        34157,
        2289,
        20182,
        18637,
        105,
        2412,
        1546,
        4715,
        40,
        12759,
        1543,
        51215
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 56,
      "seek": 17740,
      "start": 194.4,
      "end": 196.4,
      "text": "从这个DeepSick后台的这个成本分析里面。",
      "tokens": [
        51215,
        35630,
        15368,
        11089,
        595,
        50,
        618,
        13547,
        15433,
        1546,
        15368,
        11336,
        8802,
        6627,
        7360,
        238,
        15759,
        8833,
        1543,
        51315
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 57,
      "seek": 17740,
      "start": 196.4,
      "end": 197.4,
      "text": "我们可以看到啊。",
      "tokens": [
        51315,
        15003,
        6723,
        18032,
        4905,
        1543,
        51365
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 58,
      "seek": 17740,
      "start": 197.4,
      "end": 199.4,
      "text": "Pi是使用的成本是最高的。",
      "tokens": [
        51365,
        47,
        72,
        1541,
        22982,
        9254,
        1546,
        11336,
        8802,
        1541,
        8661,
        12979,
        1546,
        1543,
        51465
      ],
      "temperature": 0,
      "avg_logprob": -0.1523422094491812,
      "compression_ratio": 1.2861736334405145,
      "no_speech_prob": 1.4411220480847753e-11,
      "source": "whisper"
    },
    {
      "id": 59,
      "seek": 19940,
      "start": 199.4,
      "end": 201.4,
      "text": "1.7亿元。",
      "tokens": [
        50365,
        16,
        13,
        22,
        1369,
        123,
        14812,
        1543,
        50465
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 60,
      "seek": 19940,
      "start": 201.4,
      "end": 202.4,
      "text": "其次是这个CodeS App。",
      "tokens": [
        50465,
        9572,
        9487,
        1541,
        15368,
        34,
        1429,
        50,
        3132,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 61,
      "seek": 19940,
      "start": 202.4,
      "end": 204.4,
      "text": "接下来就是Cloud Code和Open Code。",
      "tokens": [
        50515,
        14468,
        4438,
        6912,
        5620,
        32787,
        15549,
        12565,
        45569,
        15549,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 62,
      "seek": 19940,
      "start": 204.4,
      "end": 206.4,
      "text": "那么这个结果非常出乎我的意料。",
      "tokens": [
        50615,
        35693,
        15368,
        45641,
        9319,
        14392,
        7781,
        2930,
        236,
        14200,
        9042,
        33404,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 63,
      "seek": 19940,
      "start": 206.4,
      "end": 208.4,
      "text": "那么我认为Pi这个框架。",
      "tokens": [
        50715,
        35693,
        1654,
        7422,
        97,
        13992,
        47,
        72,
        15368,
        21416,
        228,
        7360,
        114,
        1543,
        50815
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 64,
      "seek": 19940,
      "start": 208.4,
      "end": 211.4,
      "text": "加上DeepSick Flash应该属于又快又省钱。",
      "tokens": [
        50815,
        9990,
        5708,
        11089,
        595,
        50,
        618,
        20232,
        44297,
        44646,
        9636,
        252,
        37732,
        17047,
        10251,
        17047,
        2862,
        223,
        39623,
        1543,
        50965
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 65,
      "seek": 19940,
      "start": 211.4,
      "end": 213.4,
      "text": "但是呢从测试结果来看。",
      "tokens": [
        50965,
        11189,
        6240,
        35630,
        11038,
        233,
        5233,
        243,
        45641,
        9319,
        6912,
        4200,
        1543,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 66,
      "seek": 19940,
      "start": 213.4,
      "end": 215.4,
      "text": "Pi使用了60分钟。",
      "tokens": [
        51065,
        47,
        72,
        22982,
        9254,
        2289,
        4550,
        6627,
        50064,
        1543,
        51165
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 67,
      "seek": 19940,
      "start": 215.4,
      "end": 217.4,
      "text": "然后成本是最高的。",
      "tokens": [
        51165,
        26636,
        11336,
        8802,
        1541,
        8661,
        12979,
        1546,
        1543,
        51265
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 68,
      "seek": 19940,
      "start": 217.4,
      "end": 218.4,
      "text": "时间是最长的。",
      "tokens": [
        51265,
        44848,
        1541,
        8661,
        32271,
        1546,
        1543,
        51315
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 69,
      "seek": 19940,
      "start": 218.4,
      "end": 219.4,
      "text": "而且效果是最差的。",
      "tokens": [
        51315,
        22942,
        43076,
        9319,
        1541,
        8661,
        21679,
        1546,
        1543,
        51365
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 70,
      "seek": 19940,
      "start": 219.4,
      "end": 221.4,
      "text": "目前还不知道具体的原因是什么。",
      "tokens": [
        51365,
        39004,
        14852,
        17572,
        39806,
        29485,
        1546,
        42577,
        1541,
        10440,
        1543,
        51465
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 71,
      "seek": 19940,
      "start": 221.4,
      "end": 223.4,
      "text": "然后CodeS App也是使用了比较长的时间。",
      "tokens": [
        51465,
        26636,
        34,
        1429,
        50,
        3132,
        22021,
        22982,
        9254,
        2289,
        11706,
        9830,
        225,
        32271,
        1546,
        44848,
        1543,
        51565
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 72,
      "seek": 19940,
      "start": 223.4,
      "end": 225.4,
      "text": "而且效果也非常差。",
      "tokens": [
        51565,
        22942,
        43076,
        9319,
        6404,
        14392,
        21679,
        1543,
        51665
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 73,
      "seek": 19940,
      "start": 225.4,
      "end": 228.4,
      "text": "可能跟它的内置的验证流程是否有关系。",
      "tokens": [
        51665,
        16657,
        9678,
        45224,
        34742,
        34719,
        1546,
        49657,
        234,
        5233,
        223,
        27854,
        29649,
        1541,
        45334,
        2412,
        28053,
        25368,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.07532074986671915,
      "compression_ratio": 1.3394255874673628,
      "no_speech_prob": 1.2456362330492965e-11,
      "source": "whisper"
    },
    {
      "id": 74,
      "seek": 22840,
      "start": 228.4,
      "end": 232.4,
      "text": "因为我们其实使用的是第三方的模型来对接Class App。",
      "tokens": [
        50365,
        34627,
        15003,
        41646,
        22982,
        9254,
        24620,
        35878,
        9249,
        1546,
        41908,
        39823,
        6912,
        8713,
        14468,
        44621,
        3132,
        1543,
        50565
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 75,
      "seek": 22840,
      "start": 232.4,
      "end": 235.4,
      "text": "相当于不属于官方的一个范围。",
      "tokens": [
        50565,
        15106,
        16233,
        37732,
        1960,
        9636,
        252,
        37732,
        31929,
        9249,
        1546,
        20182,
        27395,
        225,
        3919,
        112,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 76,
      "seek": 22840,
      "start": 235.4,
      "end": 237.4,
      "text": "然后表现最好的是Cloud Code。",
      "tokens": [
        50715,
        26636,
        17571,
        20204,
        8661,
        20715,
        1541,
        32787,
        15549,
        1543,
        50815
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 77,
      "seek": 22840,
      "start": 237.4,
      "end": 238.4,
      "text": "这个应该是在我的意料中的。",
      "tokens": [
        50815,
        15368,
        44297,
        44646,
        1541,
        3581,
        14200,
        9042,
        33404,
        5975,
        1546,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 78,
      "seek": 22840,
      "start": 238.4,
      "end": 240.4,
      "text": "毕竟它有那么一套成熟的一个流程。",
      "tokens": [
        50865,
        7256,
        243,
        11957,
        253,
        11284,
        2412,
        35693,
        2257,
        1881,
        245,
        11336,
        21805,
        253,
        1546,
        20182,
        27854,
        29649,
        1543,
        50965
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 79,
      "seek": 22840,
      "start": 240.4,
      "end": 244.4,
      "text": "那么我在测试过程中也是把所有的技能都全部禁用掉了。",
      "tokens": [
        50965,
        35693,
        40528,
        11038,
        233,
        5233,
        243,
        16866,
        29649,
        5975,
        22021,
        16075,
        39300,
        1546,
        32502,
        8225,
        7182,
        38714,
        18614,
        223,
        9254,
        29327,
        2289,
        1543,
        51165
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 80,
      "seek": 22840,
      "start": 244.4,
      "end": 250.4,
      "text": "然后基本上用纯内置提示词加上这个模型的能力来跑这样一个结果。",
      "tokens": [
        51165,
        26636,
        37946,
        5708,
        9254,
        16853,
        107,
        34742,
        34719,
        20949,
        25696,
        5233,
        235,
        9990,
        5708,
        15368,
        41908,
        39823,
        1546,
        8225,
        13486,
        6912,
        32585,
        21209,
        20182,
        45641,
        9319,
        1543,
        51465
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 81,
      "seek": 22840,
      "start": 250.4,
      "end": 252.4,
      "text": "所以从我这个测试结果啊。",
      "tokens": [
        51465,
        7239,
        35630,
        1654,
        15368,
        11038,
        233,
        5233,
        243,
        45641,
        9319,
        4905,
        1543,
        51565
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 82,
      "seek": 22840,
      "start": 252.4,
      "end": 254.4,
      "text": "虽然说不代表一个权威啊。",
      "tokens": [
        51565,
        12026,
        121,
        5823,
        8090,
        1960,
        19105,
        17571,
        20182,
        4422,
        225,
        19976,
        223,
        4905,
        1543,
        51665
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 83,
      "seek": 22840,
      "start": 254.4,
      "end": 257.4,
      "text": "来看还是推荐大家使用Cloud Code去对接DeepSick Flash。",
      "tokens": [
        51665,
        6912,
        4200,
        45726,
        33597,
        31409,
        238,
        6868,
        22982,
        9254,
        32787,
        15549,
        6734,
        8713,
        14468,
        11089,
        595,
        50,
        618,
        20232,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.08598965437954696,
      "compression_ratio": 1.3865979381443299,
      "no_speech_prob": 1.523139253611916e-11,
      "source": "whisper"
    },
    {
      "id": 84,
      "seek": 25740,
      "start": 257.4,
      "end": 259.4,
      "text": "会比较好一点。",
      "tokens": [
        50365,
        12949,
        11706,
        9830,
        225,
        2131,
        49042,
        1543,
        50465
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 85,
      "seek": 25740,
      "start": 259.4,
      "end": 260.4,
      "text": "毕竟生态比较丰富嘛。",
      "tokens": [
        50465,
        7256,
        243,
        11957,
        253,
        8244,
        3757,
        223,
        11706,
        9830,
        225,
        940,
        108,
        47564,
        20722,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 86,
      "seek": 25740,
      "start": 260.4,
      "end": 263.4,
      "text": "然后或者说你使用Cloud X也行。",
      "tokens": [
        50515,
        26636,
        31148,
        8090,
        2166,
        22982,
        9254,
        32787,
        1783,
        6404,
        8082,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 87,
      "seek": 25740,
      "start": 263.4,
      "end": 265.4,
      "text": "那么你可以推荐你使用这个Open Code。",
      "tokens": [
        50665,
        35693,
        42766,
        33597,
        31409,
        238,
        2166,
        22982,
        9254,
        15368,
        45569,
        15549,
        1543,
        50765
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 88,
      "seek": 25740,
      "start": 265.4,
      "end": 268.4,
      "text": "那么其实不管是Cloud Code还是Cloud App。",
      "tokens": [
        50765,
        35693,
        41646,
        45376,
        1541,
        32787,
        15549,
        45726,
        32787,
        3132,
        1543,
        50915
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 89,
      "seek": 25740,
      "start": 268.4,
      "end": 270.4,
      "text": "他们的内置的提示词或者流程,",
      "tokens": [
        50915,
        47911,
        1546,
        34742,
        34719,
        1546,
        20949,
        25696,
        5233,
        235,
        31148,
        27854,
        29649,
        11,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 90,
      "seek": 25740,
      "start": 270.4,
      "end": 272.4,
      "text": "肯定是优先适配于自己的模型。",
      "tokens": [
        51015,
        14356,
        107,
        12088,
        1541,
        7384,
        246,
        10108,
        2215,
        224,
        38846,
        37732,
        17645,
        1546,
        41908,
        39823,
        1543,
        51115
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 91,
      "seek": 25740,
      "start": 272.4,
      "end": 275.4,
      "text": "那比如说在之前的视频里面我分享过了。",
      "tokens": [
        51115,
        4184,
        36757,
        8090,
        3581,
        32442,
        1546,
        40656,
        39752,
        15759,
        8833,
        1654,
        30855,
        16866,
        2289,
        1543,
        51265
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 92,
      "seek": 25740,
      "start": 275.4,
      "end": 279.4,
      "text": "Cloud Code在OPUS5.0之后会减省80%的内置提示词。",
      "tokens": [
        51265,
        32787,
        15549,
        3581,
        12059,
        3447,
        20,
        13,
        15,
        9574,
        13547,
        12949,
        6336,
        237,
        2862,
        223,
        4702,
        4,
        1546,
        34742,
        34719,
        20949,
        25696,
        5233,
        235,
        1543,
        51465
      ],
      "temperature": 0,
      "avg_logprob": -0.14382916026645237,
      "compression_ratio": 1.2640264026402641,
      "no_speech_prob": 1.602515342646882e-11,
      "source": "whisper"
    },
    {
      "id": 93,
      "seek": 27940,
      "start": 279.4,
      "end": 282.4,
      "text": "那么针对非OPUS5.0的模型。",
      "tokens": [
        50365,
        35693,
        25153,
        230,
        8713,
        12107,
        12059,
        3447,
        20,
        13,
        15,
        1546,
        41908,
        39823,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 94,
      "seek": 27940,
      "start": 282.4,
      "end": 285.4,
      "text": "他们会用其他的这种提示词。",
      "tokens": [
        50515,
        47911,
        12949,
        9254,
        9572,
        31309,
        5562,
        39810,
        20949,
        25696,
        5233,
        235,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 95,
      "seek": 27940,
      "start": 285.4,
      "end": 289.4,
      "text": "那么可能当DeepSick的能力超过OPUS5.0。",
      "tokens": [
        50665,
        35693,
        16657,
        16233,
        11089,
        595,
        50,
        618,
        1546,
        8225,
        13486,
        19869,
        16866,
        12059,
        3447,
        20,
        13,
        15,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 96,
      "seek": 27940,
      "start": 289.4,
      "end": 291.4,
      "text": "那么再使用Cloud Code的时候。",
      "tokens": [
        50865,
        35693,
        8623,
        22982,
        9254,
        32787,
        15549,
        49873,
        1543,
        50965
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 97,
      "seek": 27940,
      "start": 291.4,
      "end": 294.4,
      "text": "那么就会用老的提示词用到乔模型上。",
      "tokens": [
        50965,
        35693,
        3111,
        12949,
        9254,
        10439,
        1546,
        20949,
        25696,
        5233,
        235,
        9254,
        4511,
        2930,
        242,
        41908,
        39823,
        5708,
        1543,
        51115
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 98,
      "seek": 27940,
      "start": 294.4,
      "end": 295.4,
      "text": "那效果可能会打折。",
      "tokens": [
        51115,
        4184,
        43076,
        9319,
        16657,
        12949,
        12467,
        7235,
        246,
        1543,
        51165
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 99,
      "seek": 27940,
      "start": 295.4,
      "end": 299.4,
      "text": "所以最好的还是模型公司自己出了这个AI变成工具。",
      "tokens": [
        51165,
        7239,
        8661,
        20715,
        45726,
        41908,
        39823,
        13545,
        32981,
        17645,
        7781,
        2289,
        15368,
        48698,
        2129,
        246,
        11336,
        23323,
        39806,
        1543,
        51365
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 100,
      "seek": 27940,
      "start": 299.4,
      "end": 300.4,
      "text": "或者我们叫Hannis工具。",
      "tokens": [
        51365,
        31148,
        15003,
        19855,
        39,
        969,
        271,
        23323,
        39806,
        1543,
        51415
      ],
      "temperature": 0,
      "avg_logprob": -0.12066665649414063,
      "compression_ratio": 1.250936329588015,
      "no_speech_prob": 1.6628384433281518e-11,
      "source": "whisper"
    },
    {
      "id": 101,
      "seek": 30040,
      "start": 300.4,
      "end": 303.4,
      "text": "DepSick官方最近也在内测自己的这个Hannis工具。",
      "tokens": [
        50365,
        35,
        595,
        50,
        618,
        31929,
        9249,
        46777,
        6404,
        3581,
        34742,
        11038,
        233,
        17645,
        1546,
        15368,
        39,
        969,
        271,
        23323,
        39806,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 102,
      "seek": 30040,
      "start": 303.4,
      "end": 305.4,
      "text": "也非常期待它的正式上线。",
      "tokens": [
        50515,
        6404,
        14392,
        16786,
        18390,
        45224,
        15789,
        27584,
        5708,
        16853,
        123,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 103,
      "seek": 30040,
      "start": 305.4,
      "end": 307.4,
      "text": "到时候我们再来测试一下。",
      "tokens": [
        50615,
        4511,
        29111,
        15003,
        8623,
        6912,
        11038,
        233,
        5233,
        243,
        8861,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 104,
      "seek": 30040,
      "start": 307.4,
      "end": 308.4,
      "text": "然后这几个工具的话。",
      "tokens": [
        50715,
        26636,
        5562,
        6336,
        254,
        7549,
        23323,
        39806,
        44575,
        1543,
        50765
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 105,
      "seek": 30040,
      "start": 308.4,
      "end": 310.4,
      "text": "如果你喜欢自己折腾。",
      "tokens": [
        50765,
        45669,
        41606,
        17645,
        7235,
        246,
        21184,
        122,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 106,
      "seek": 30040,
      "start": 310.4,
      "end": 312.4,
      "text": "那希望自己去配置单独的包。",
      "tokens": [
        50865,
        4184,
        29955,
        17645,
        6734,
        38846,
        34719,
        47446,
        18637,
        105,
        1546,
        23305,
        1543,
        50965
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 107,
      "seek": 30040,
      "start": 312.4,
      "end": 314.4,
      "text": "或者说第三方的扩展。",
      "tokens": [
        50965,
        31148,
        8090,
        35878,
        9249,
        1546,
        3416,
        102,
        43491,
        1543,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 108,
      "seek": 30040,
      "start": 314.4,
      "end": 316.4,
      "text": "那么我非常推荐使用Pi。",
      "tokens": [
        51065,
        35693,
        1654,
        14392,
        33597,
        31409,
        238,
        22982,
        9254,
        47,
        72,
        1543,
        51165
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 109,
      "seek": 30040,
      "start": 316.4,
      "end": 319.4,
      "text": "因为Pi其实是一个非常简单的一个Agent工具。",
      "tokens": [
        51165,
        34627,
        47,
        72,
        41646,
        1541,
        20182,
        14392,
        11249,
        222,
        47446,
        1546,
        20182,
        32,
        6930,
        23323,
        39806,
        1543,
        51315
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 110,
      "seek": 30040,
      "start": 319.4,
      "end": 321.4,
      "text": "它不仅可以开发自己的Agent工具。",
      "tokens": [
        51315,
        11284,
        1960,
        1550,
        227,
        6723,
        18937,
        28926,
        17645,
        1546,
        32,
        6930,
        23323,
        39806,
        1543,
        51415
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 111,
      "seek": 30040,
      "start": 321.4,
      "end": 323.4,
      "text": "使用它来进行编程。",
      "tokens": [
        51415,
        22982,
        9254,
        11284,
        6912,
        36700,
        8082,
        38109,
        244,
        29649,
        1543,
        51515
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 112,
      "seek": 30040,
      "start": 323.4,
      "end": 325.4,
      "text": "当然它的官网也有非常的第三方包。",
      "tokens": [
        51515,
        40486,
        45224,
        31929,
        16469,
        239,
        6404,
        2412,
        48263,
        35878,
        9249,
        23305,
        1543,
        51615
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 113,
      "seek": 30040,
      "start": 325.4,
      "end": 329.4,
      "text": "比如说像子代理,像动态工作流,像目标。",
      "tokens": [
        51615,
        36757,
        8090,
        12760,
        7626,
        19105,
        13876,
        11,
        12760,
        34961,
        3757,
        223,
        41315,
        27854,
        11,
        12760,
        11386,
        162,
        3921,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.10537213087081909,
      "compression_ratio": 1.389920424403183,
      "no_speech_prob": 1.5910132586394177e-11,
      "source": "whisper"
    },
    {
      "id": 114,
      "seek": 32940,
      "start": 329.4,
      "end": 332.4,
      "text": "以及其他的非常好用的一些包。",
      "tokens": [
        50365,
        40282,
        9572,
        31309,
        14392,
        2131,
        9254,
        1546,
        38515,
        23305,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.1418332271888608,
      "compression_ratio": 1.24,
      "no_speech_prob": 1.4174096792107793e-11,
      "source": "whisper"
    },
    {
      "id": 115,
      "seek": 32940,
      "start": 332.4,
      "end": 334.4,
      "text": "那你把它按需要去加载进来。",
      "tokens": [
        50515,
        4184,
        2166,
        42061,
        26613,
        35748,
        6734,
        9990,
        17819,
        121,
        36700,
        6912,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.1418332271888608,
      "compression_ratio": 1.24,
      "no_speech_prob": 1.4174096792107793e-11,
      "source": "whisper"
    },
    {
      "id": 116,
      "seek": 32940,
      "start": 334.4,
      "end": 336.4,
      "text": "那也是一个非常好用的工具。",
      "tokens": [
        50615,
        4184,
        22021,
        20182,
        14392,
        2131,
        9254,
        1546,
        23323,
        39806,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.1418332271888608,
      "compression_ratio": 1.24,
      "no_speech_prob": 1.4174096792107793e-11,
      "source": "whisper"
    },
    {
      "id": 117,
      "seek": 32940,
      "start": 336.4,
      "end": 337.4,
      "text": "那么视频就到这儿。",
      "tokens": [
        50715,
        35693,
        40656,
        39752,
        45918,
        5562,
        22933,
        1543,
        50765
      ],
      "temperature": 0,
      "avg_logprob": -0.1418332271888608,
      "compression_ratio": 1.24,
      "no_speech_prob": 1.4174096792107793e-11,
      "source": "whisper"
    },
    {
      "id": 118,
      "seek": 32940,
      "start": 337.4,
      "end": 338.4,
      "text": "希望这个视频对你有所帮助。",
      "tokens": [
        50765,
        29955,
        15368,
        40656,
        39752,
        8713,
        43320,
        5966,
        4845,
        106,
        37618,
        1543,
        50815
      ],
      "temperature": 0,
      "avg_logprob": -0.1418332271888608,
      "compression_ratio": 1.24,
      "no_speech_prob": 1.4174096792107793e-11,
      "source": "whisper"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 339.081,
    "reference_path": null,
    "reference_kind": null,
    "reference_language": null,
    "selected_source": "whisper_segmented",
    "decision_reason": "source_unavailable_or_untrusted_whisper_passed",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 0,
      "duration_coverage": 0.0,
      "usable_full": false,
      "usable_partial": false,
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "language": {
        "requested": "zh",
        "matched": false
      }
    },
    "whisper_profile": {
      "segment_count": 119,
      "first_start": 0.0,
      "final_end": 338.4,
      "duration_coverage": 0.997992,
      "maximum_gap_seconds": 0.0,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 417,
        "han": 1582,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "comparison": {
      "window_seconds": 30.0,
      "compared_window_count": 0,
      "median_score": 0.0,
      "divergent_window_count": 0,
      "divergent_ratio": 0.0,
      "divergent_windows": []
    }
  }
}