智能体
x.com ·
@_philschmid
Gemini 3.8 Live 与 3.8 Live Extended Thinking 发布
Gemini 3.8 Live 与 3.8 Live Extended Thinking 发布,延续对实时语音 Agent 的投入。官方给出的指标包括 Artificial Analysis 质量指数 82.6(第一)、输入 $0.005/分钟与输出 $0.018/分钟、τ-banking 智能体任务完成 35.1(第一)。Extended Thinking 可在后台思考并异步执行工具调用,在不打断对话的情况下协调多步工具使用。已在 Google AI Studio、Gemini API、Google Search、Gemini App 上线,并作为 LiveKit、Pipecat、LangChain、Vercel 的合作伙伴插件提供。
AI推荐理由 Philipp Schmid 发布 Gemini 3.8 Live 与 3.8 Live Extended Thinking,给出 Artificial Analysis 质量指数 82.6(第一)、输入 $0.005/分钟与输出 $0.018/分钟、τ-banking 智能体任务完成 35.1(第一)等具体指标,并说明 Extended Thinking 可在后台思考并异步执行工具调用,同时列出 Google AI Studio、Gemini API、Google Search、Gemini App 及 LiveKit、Pipecat、LangChain、Vercel 等接入渠道。属于模型发布的一手信息,指标与定价明确,对实时语音 Agent 开发者价值高;但数据来自发布方口径,需以官方文档核实。
82
浏览 4,244
点赞 79
评论 5
转发 7
正文
Gemini 3.8 Live and 3.8 Live Extended Thinking are here. Following 3.5 Transcribe last month, this continues our focus on real-time voice agents. 🐸🐸
- 82.6 (#1) on Artificial Analysis Quality Index
- $0.005/min input and $0.018/min output.
- 35.1 (#1) Agentic task completion (τ-banking)
3.8 Live Extended Thinking can think in the background and run asynchronously tool calls while keep talking to you. It uses background thinking to coordinate multi-step tool use without losing conversational flow.
Available in [@GoogleAIStudio](https://x.com/GoogleAIStudio) Gemini API, Google Search, and [@GeminiApp](https://x.com/GeminiApp) or as partner plugins in [@LiveKit](https://x.com/LiveKit), [@Pipecat\_ai](https://x.com/Pipecat_ai), [@LangChainAI](https://x.com/LangChainAI), and [@vercel](https://x.com/vercel)
[Your browser does not support the video tag.](https://video.twimg.com/amplify_video/2099907963431378944/vid/avc1/1920x1080/TIOKa2kwFrK-uCTN.mp4?tag=29)
[🔗 View on Twitter](https://x.com/_philschmid/status/2099908172899357093/video/1)