Claude Agent SDK

What is Claude Agent SDK?

The Claude Agent SDK for Python lets developers embed the Claude Code agent loop inside their own programs. You install it with pip, and the Claude Code CLI comes bundled in the package, so nothing else needs installing on Python 3.10 or newer. A simple query() call streams back messages from Claude, which by default can use the full Claude Code toolset, including Read, Write, Edit and Bash. For richer apps, ClaudeSDKClient supports two-way conversations, custom tools written as plain Python functions and served as in-process MCP servers, and hooks that run your code at points in the agent loop, such as blocking a risky shell command. Options control the working directory, system prompt, allowed tools and permission mode. The SDK is distributed free on PyPI and GitHub under Anthropic's Commercial Terms, and no separate price is listed, because running it requires access to Claude.

Use cases

• Build a coding agent in Python
• Automate file and shell tasks with Claude
• Add custom tools as in-process MCP servers
• Block risky commands with permission hooks
• Run multi-turn agent conversations from code
• Embed Claude Code in internal developer tools

Who it's for

• Python developers
• AI engineers
• Platform teams

Why use Claude Agent SDK

• The same agent loop that powers Claude Code runs inside your own application, so you skip building tool handling yourself.
• Bundling the CLI means installation is a single pip command.
• In-process custom tools avoid managing separate server processes and are easier to debug.
• Hooks give deterministic control over what the agent may run.
• Permission options let you decide which tool calls are auto-approved.

Claude Agent SDK features

• query() function that streams messages from the Claude agent loop
• ClaudeSDKClient for interactive, bidirectional conversations
• Full Claude Code toolset available by default, including Read, Write, Edit and Bash
• Custom tools written as Python functions with the @tool decorator
• In-process SDK MCP servers, mixable with external MCP servers
• Hooks such as PreToolUse to approve, deny or adjust tool calls
• ClaudeAgentOptions for system prompt, working directory and permission mode
• allowed_tools and disallowed_tools to control approvals
• Typed error classes such as CLINotFoundError and ProcessError
• Claude Code CLI bundled in the package, with an option to use your own

Claude Agent SDK integrations

• Claude Code
• MCP
• PyPI
• GitHub Actions

How to get started with Claude Agent SDK

1. Install the package with pip install claude-agent-sdk on Python 3.10 or newer.
2. Call query() with a prompt, or open a ClaudeSDKClient for a back-and-forth session.
3. Set ClaudeAgentOptions for the working directory, system prompt, allowed tools and permission mode.
4. Add custom tools or hooks as Python functions so Claude can call your code and you can govern its actions.
5. Handle the streamed messages and catch SDK errors in your application.

Claude Agent SDK limitations

• Requires Python 3.10 or later
• Using it requires access to Claude, which is covered by Anthropic's own terms
• allowed_tools only pre-approves tools; it does not remove them from Claude's toolset
• Changing a custom system prompt has no effect on an existing session unless snapshotting is turned off

Claude Agent SDK pricing

We couldn't confirm pricing on the official site. Check github.com for current prices.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

Claude Agent SDK alternatives

The main Claude Agent SDK alternatives listed on Debuly are OpenAI Agents SDK, LangGraph, CrewAI, Pydantic AI, AutoGen and LangChain Deep Agents. The table compares pricing and platforms; order is A–Z.

1. 

   AutoGen

   Microsoft's MIT-licensed Python and .NET framework for multi-agent AI apps, now archived; new projects are pointed to Microsoft Agent Framework.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit AutoGen (official site)
2. 

   CrewAI

   Multi-agent framework and platform for building AI agent crews in Python or a visual editor, then deploying and monitoring them.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit CrewAI (official site)
3. 

   LangChain Deep Agents

   Open-source LangChain agent harness: long-term memory, a virtual filesystem, subagents and planning included

   Developer platformsAutomation and agentsCoding
   Open sourceVisit LangChain Deep Agents (official site)
4. 

   LangGraph

   Low-level open-source orchestration framework and runtime for building stateful, long-running AI agents as graphs.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit LangGraph (official site)
5. 

   OpenAI Agents SDK

   Open-source Python framework for building multi-agent workflows with tools, handoffs, guardrails, tracing, sandboxes and voice agents.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit OpenAI Agents SDK (official site)
6. 

   Pydantic AI

   A type-safe, model-agnostic Python framework for building agents and LLM applications, from the team behind Pydantic.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit Pydantic AI (official site)

Claude Agent SDK FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from avatars.githubusercontent.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.