{
  "version": 1,
  "event_id": "evt_22c066d4aebc48ea",
  "url": "https://xiyu.news/events/evt_22c066d4aebc48ea/",
  "json": "https://xiyu.news/api/events/evt_22c066d4aebc48ea.json",
  "type": "other",
  "status": "monitoring",
  "category": "technology",
  "title": {
    "zh": "Mistral发布Voxtral 3：把语音识别、推理和工具调用接成实时Agent",
    "en": "Voxtral 3"
  },
  "current_state": {
    "zh": "Mistral 发布了 Voxtral 3，这是一个模块化语音栈，包含转录模型、语音合成模型和基于 LiveKit 的 Voice Agent SDK，支持 50 多种语言的转录与实时语音 Agent；该模型仍为闭源。",
    "en": "Mistral released Voxtral 3, a modular speech stack with transcription models, a synthesis model, and a LiveKit-based Voice Agent SDK, supporting 50+ languages for transcription and real-time voice agents; the model remains closed-source."
  },
  "first_seen_at": "2026-10-20T17:25:49.327375+00:00",
  "last_updated_at": "2026-10-21T06:41:28.555697+00:00",
  "last_material_change_at": "2026-10-21T06:41:28.555697+00:00",
  "confidence": 0.75,
  "updates_count": 2,
  "sources_count": 2,
  "entities": [
    "agent",
    "mistral",
    "voxtral",
    "voxtral-3"
  ],
  "identifiers": [],
  "topics": [
    "ai-models",
    "closed-source",
    "mistral",
    "speech-recognition",
    "text-to-speech",
    "voice-ai",
    "voxtral-3"
  ],
  "updates": [
    {
      "update_id": "upd_733f31f91bfd9dc2",
      "event_id": "evt_22c066d4aebc48ea",
      "occurred_at": "2026-10-20T16:02:07Z",
      "published_at": "2026-10-20T16:02:07Z",
      "first_seen_at": "2026-10-20T17:25:49.327375Z",
      "time_precision": "published",
      "update_type": "initial",
      "material_change": true,
      "title_zh": "Voxtral 3",
      "title_en": "Voxtral 3",
      "what_changed_zh": "Mistral 发布 Voxtral 3，这是一款升级版语音模型，提供转写、说话人分离和语音生成能力，但仍为闭源。",
      "what_changed_en": "Mistral released Voxtral 3, an upgraded speech model offering transcription, diarization, and speech-generation capabilities, though it remains closed-source.",
      "current_state_zh": "Mistral 发布 Voxtral 3，这是一款升级版语音模型，提供转写、说话人分离和语音生成能力，但仍为闭源。",
      "current_state_en": "Mistral released Voxtral 3, an upgraded speech model offering transcription, diarization, and speech-generation capabilities, though it remains closed-source.",
      "detailed_summary_zh": "Mistral released Voxtral 3, an upgraded speech model offering transcription, diarization, and speech-generation capabilities, though it remains closed-source.",
      "detailed_summary_en": "Mistral released Voxtral 3, an upgraded speech model offering transcription, diarization, and speech-generation capabilities, though it remains closed-source.",
      "background_zh": "",
      "background_en": "",
      "community_discussion_zh": "",
      "community_discussion_en": "",
      "market_impact_zh": "",
      "market_impact_en": "",
      "importance_score": 7.0,
      "references": [
        {
          "url": "https://news.ycombinator.com/item?id=50168396",
          "title": "Community discussion"
        }
      ],
      "confidence": 0.75,
      "story_ids": [
        "hackernews:story:50168396"
      ],
      "sources": [
        {
          "url": "https://mistral.ai/news/voxtral-3/",
          "label": "ilreb",
          "source_type": "hackernews",
          "official": false
        }
      ]
    },
    {
      "update_id": "upd_c0c91bd4f47cd277",
      "event_id": "evt_22c066d4aebc48ea",
      "occurred_at": "2026-10-21T03:35:02Z",
      "published_at": "2026-10-21T03:35:02Z",
      "first_seen_at": "2026-10-21T06:41:28.555697Z",
      "time_precision": "published",
      "update_type": "confirmation",
      "material_change": true,
      "title_zh": "Mistral发布Voxtral 3：把语音识别、推理和工具调用接成实时Agent",
      "title_en": "Mistral发布Voxtral 3：把语音识别、推理和工具调用接成实时Agent",
      "what_changed_zh": "新增披露 Voxtral 3 为模块化语音栈，包含三款转录模型、一款语音合成模型和基于 LiveKit 的 Voice Agent SDK；转录支持 50 多种语言，语音合成支持 35 种以上语言和 50 多种方言，并公布 API 定价。",
      "what_changed_en": "Added details that Voxtral 3 is a modular speech stack with three transcription models, a synthesis model, and a LiveKit-based Voice Agent SDK; transcription supports 50+ languages, synthesis supports 35+ languages and 50+ dialects, with API pricing disclosed.",
      "current_state_zh": "Mistral 发布了 Voxtral 3，这是一个模块化语音栈，包含转录模型、语音合成模型和基于 LiveKit 的 Voice Agent SDK，支持 50 多种语言的转录与实时语音 Agent；该模型仍为闭源。",
      "current_state_en": "Mistral released Voxtral 3, a modular speech stack with transcription models, a synthesis model, and a LiveKit-based Voice Agent SDK, supporting 50+ languages for transcription and real-time voice agents; the model remains closed-source.",
      "detailed_summary_zh": "Mistral released Voxtral 3, a modular speech-to-speech stack with transcription and synthesis models plus a LiveKit-based Voice Agent SDK for real-time tool-calling agents.",
      "detailed_summary_en": "Mistral released Voxtral 3, a modular speech-to-speech stack with transcription and synthesis models plus a LiveKit-based Voice Agent SDK for real-time tool-calling agents.",
      "background_zh": "",
      "background_en": "",
      "community_discussion_zh": "",
      "community_discussion_en": "",
      "market_impact_zh": "",
      "market_impact_en": "",
      "importance_score": 7.0,
      "references": [],
      "confidence": 0.95,
      "story_ids": [
        "telegram:theblockbeats:201544"
      ],
      "sources": [
        {
          "url": "https://m.theblockbeats.info/flash/372884?from=telegram",
          "label": "theblockbeats",
          "source_type": "telegram",
          "official": false
        }
      ]
    }
  ]
}