---
title: "llm-anthropic 0.26 发布：新增 Claude 5 系列模型与服务器端工具支持"
source: "Simon Willison 博客（Atom）"
source_tier: "t2"
url: "https://simonwillison.net/2026/Aug/4/llm-anthropic/#atom-everything"
permalink: "https://feed.astralor.com/items/itm_dace681f7641d31c02df7d66"
published: "2026-08-04T22:00:58.000Z"
score: 73
channel: "ai"
tags: ["Agent 工具","Anthropic","Claude 5","llm-anthropic","LLM CLI","MCP"]
---

# llm-anthropic 0.26 发布：新增 Claude 5 系列模型与服务器端工具支持

## AI 点评

> 这次更新最值得关注的是服务器端工具成为一等公民，表明 Anthropic 正在把 Agent 能力下沉到 API 和 CLI 层。对于做自动化流程或复杂工具链的开发者，Claude 5 的默认思考机制会改变响应延迟和输出模式，最好提前测试适配。

## 速览

llm-anthropic 0.26 发布，依托 LLM 0.32 新增 Claude Fable 5、Sonnet 5 和 Opus 5 模型，并支持 WebSearch、WebFetch、CodeExecution 和 AnthropicMCP 等服务器端工具，可通过 LLM 的 -T 接口或 Python tools= 调用。旧的 -o web_search* 选项被移除，改用 -T WebSearch。推理过程和工具调用结果现在以类型化事件流式传输，CLI 的推理默认输出到标准错误，可用 -R 隐藏。Claude 5 模型默认思考，提供 thinking_effort 参数控制强度，Fable 5 始终思考。整体提升了模型集成体验，简化了扩展思考配置。

## 关键点

- 新增 claude-fable-5、claude-sonnet-5、claude-opus-5 三个模型。
- 新增 WebSearch、WebFetch、CodeExecution、AnthropicMCP 服务器端工具，通过 -T 或 Python tools= 使用。
- 推理、工具调用和结果以类型化事件流式传输；CLI 推理默认输出到 stderr，可用 -R 隐藏。
- Claude 5 模型默认思考，支持 thinking_effort 低/中/高/超高/最大，Fable 5 始终思考。

## 全文

Release: llm-anthropic 0.26 Includes new features enabled by LLM 0.32: New models: claude-fable-5, claude-sonnet-5, and claude-opus-5. #75, #76 Added server-side tools for WebSearch, WebFetch, CodeExecution, and AnthropicMCP, available through LLM's -T interface or Python tools=. The previous -o web_search* options have been removed in favor of -T WebSearch. #79 Upgraded to llm>=0.32. Reasoning, tool calls, tool results, and server-side tool results now stream as typed events. Reasoning for llm CLI prompts now displays to standard error unless you pass --hide-reasoning/-R. Simplified extended thinking to thinking and thinking_effort (low, medium, high, xhigh, or max). Claude 5 models think by default; -o thinking 0 disables thinking for Sonnet 5 and Opus 5, while Fable 5 always thinks. -R/--hide-reasoning now omits reasoning from responses and logs. The thinking_budget, thinking_display, and thinking_adaptive options have been removed. #80 Tags: llm, anthropic, claude, model-context-protocol

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_dace681f7641d31c02df7d66