AI Agent周报：GPT-6.1与Gemini 4发布
From more efficient computer-use models and open weights to multi-model agent pl…
From more efficient computer-use models and open weights to multi-model agent platforms and frontier safety, the period since our last edition brought major developments across the AI-agent stack.
从更高效的计算机使用模型和开放权重，到多模型智能体平台和前沿安全领域，自上一期以来，AI 智能体栈带来了重大进展。
Here’s everything worth catching up on in the latest edition of AI Agents Pulse:
以下是本期《AI 智能体脉搏》中值得关注的全部内容：
1️⃣ OpenAI released GPT-6.1 Terra for visual reasoning, computer use and writing in ChatGPT Work, Codex and the API. Standard API pricing is $1 per million input tokens and $5 per million output tokens. https://openai.com/index/introducing-gpt-6-1-terra/
1️⃣ OpenAI 发布了 GPT-6.1 Terra，用于 ChatGPT Work、Codex 和 API 中的视觉推理、计算机使用和写作。标准 API 定价为每百万输入令牌 1 美元，每百万输出令牌 5 美元。 https://openai.com/index/introducing-gpt-6-1-terra/
2️⃣ Google announced Gemini 4 Carbon for coding, cybersecurity and long-horizon tasks. Access is starting with a subset of cyber defenders through the Fairwind Program, ahead of broader developer, business and consumer availability. https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
2️⃣ Google 宣布了 Gemini 4 Carbon，适用于编码、网络安全和长周期任务。访问权限首先通过 Fairwind 计划向部分网络防御者开放，随后将面向更广泛的开发者、企业和消费者。 https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
3️⃣ Mistral released improved Mistral Large 4 weights for agentic coding and defensive cybersecurity, with Low, Medium and High reasoning modes. Its modified MIT license requires a separate commercial license for businesses operating a Model-as-a-Service offering. https://mistral.ai/news/mistral-large-4-oss/
3️⃣ Mistral 发布了改进版的 Mistral Large 4 权重，支持智能体编码和防御性网络安全，并提供低、中、高三种推理模式。其修改后的 MIT 许可证要求提供“模型即服务”（Model-as-a-Service）业务的企业单独获取商业许可证。 https://mistral.ai/news/mistral-large-4-oss/
4️⃣ Liquid AI released the first LFM3 models in the 2B-parameter range, bringing text, vision, audio, omni and decision models onto a shared backbone. Base and post-trained weights are available on Hugging Face, alongside Playground and API access. https://www.liquid.ai/blog/lfm3
4️⃣ Liquid AI 发布了首批 LFM3 模型，参数量在 2B 级别，将文本、视觉、音频、全能型和决策型模型整合到共享主干上。基础权重和经过后训练的权重已在 Hugging Face 上提供，同时支持 Playground 和 API 访问。 https://www.liquid.ai/blog/lfm3
5️⃣ Google announced Genie 4, a world model for real-time interactive simulation with native audio. Access began with limited trusted testers; Project Genie access for Google AI Ultra subscribers and Gemini API early access are still coming. https://blog.google/innovation-and-ai/models-and-research/google-deepmind/introducing-genie-4/
5️⃣ Google 宣布了 Genie 4，这是一个支持原生音频的实时交互式模拟世界模型。访问权限最初仅限于少量受信任的测试人员；Google AI Ultra 订阅用户的 Project Genie 访问权限以及 Gemini API 的早期访问权限仍在陆续推出中。 https://blog.google/innovation-and-ai/models-and-research/google-deepmind/introducing-genie-4/
6️⃣ GitHub made Agentic Workflows generally available. Developers define automation in Markdown, which compiles to GitHub Actions YAML, using Copilot or another coding agent. Agents run with read-only permissions by default, with proposed changes checked before application. https://github.blog/changelog/2026-10-28-github-agentic-workflows-generally-available/
6️⃣ GitHub 使 Agentic Workflows 正式全面可用。开发者使用 Markdown 定义自动化流程，并通过 Copilot 或其他编码代理将其编译为 GitHub Actions YAML。默认情况下，代理以只读权限运行，并在应用前检查拟议更改。 https://github.blog/changelog/2026-10-28-github-agentic-workflows-generally-available/
7️⃣ Microsoft introduced the Copilot harness in Foundry Agent Service in public preview. It manages the agent loop and sandbox for long-running tasks, while developers supply instructions, skills, tools and model choice. https://devblogs.microsoft.com/foundry/microsoft-copilot-harness-in-foundry/
7️⃣ Microsoft 在 Foundry Agent Service 的公开预览版中引入了 Copilot harness。它负责管理长期任务的智能体循环和沙箱环境，而开发者则提供指令、技能、工具和模型选择。 https://devblogs.microsoft.com/foundry/microsoft-copilot-harness-in-foundry/
8️⃣ GitHub introduced the Copilot app and Copilot Fusion at Universe. The app connects agents and tools across projects, while Fusion chooses how to execute a task across one or more models. https://github.blog/news-insights/company-news/github-universe-2026-evolving-every-layer-of-the-stack/
8️⃣ GitHub 在 Universe 大会上推出了 Copilot 应用和 Copilot Fusion。该应用连接跨项目的智能体与工具，而 Fusion 则决定如何在一个或多个模型上执行任务。 https://github.blog/news-insights/company-news/github-universe-2026-evolving-every-layer-of-the-stack/
9️⃣ OpenAI released GPT-6 Rosalind for multistep life-sciences research with scientific tool use. Access is limited to organizations approved for Rosalind Research and their authorized users, rather than general API availability. https://developers.openai.com/api/docs/changelog
9️⃣ OpenAI 发布了 GPT-6 Rosalind，用于支持多步骤生命科学研究的科学工具使用。访问权限仅限于获得 Rosalind Research 批准的组织及其授权用户，而非面向通用 API 开放。 https://developers.openai.com/api/docs/changelog
🔟 OpenAI updated its Model Spec with guidance on context, oversight and human agency, organized around five values. The company said its models do not yet fully reflect the intended behavior described in the Spec. https://openai.com/index/updated-model-spec/
🔟 OpenAI 更新了其《模型规范》（Model Spec），围绕五项核心价值观提供了关于上下文、监督及人类能动性的指导。该公司表示，其模型尚未完全体现《规范》中描述的预期行为。 https://openai.com/index/updated-model-spec/
1️⃣1️⃣ OpenAI detailed expanded monitoring for frontier agents across training, evaluation and internal use, with adaptations for external deployment. Its layered system combines activation probes, model reviewers, investigative agents and human response to flag unauthorized or deceptive behavior. https://openai.com/index/how-we-detect-frontier-model-misalignment/
1️⃣1️⃣ OpenAI 详细说明了针对前沿智能体在训练、评估和内部使用阶段的扩展监控措施，并针对外部部署进行了适配。其分层系统结合了激活探针、模型审查员、调查性智能体和人工响应，以标记未经授权或欺骗性行为。 https://openai.com/index/how-we-detect-frontier-model-misalignment/
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力