Messages API
Send conversational requests to Claude through `POST /v1/messages`, with support for the core message-based integration pattern.
Claude Platform
Freemium
Claude Platform gives developers programmatic access to Claude models and managed agent infrastructure for building agents and applications. It includes a REST API, official client SDKs, a web Console, and documentation for direct integrations.
什么是 Claude Platform？
Claude Platform is a developer platform for building applications and agents with Claude models and managed agent infrastructure. It offers programmatic access through the RESTful Claude API at api.anthropic.com, as well as a web Console for trying requests and creating API keys.
The platform supports direct conversational requests through the Messages API and includes supporting APIs for batch processing, token counting, model discovery, file management, and custom agent skills. Beta APIs add reusable agent configurations, stateful sessions in managed cloud sandboxes, and configurable environments for agent sessions.
Developers can begin with a Claude Console account and an API key, or use Workload Identity Federation. Official client SDKs simplify integration across several languages, while the documentation and Cookbook provide implementation examples for conversations, tools, multimodal inputs, context management, and agent workflows.
Claude Platform 能做什么？
Batch and token APIs
Process large volumes of Messages requests asynchronously with the Message Batches API, and count tokens before sending requests to help manage costs and rate limits.
Files and custom skills
Upload and manage files for use across API calls, and create or manage custom agent skills through the Files API and Skills API.
Managed agent infrastructure
Use the beta Agents, Sessions, and Environments APIs to define reusable agent configurations, run stateful sessions in managed cloud sandboxes, and configure session environments.
Official client SDKs
Anthropic SDKs provide type-safe request and response handling, automatic authentication and version headers, built-in retry and error handling, streaming support, and request timeouts.
Console and authentication options
Use the web Console playground to try the API and create keys, with API-key authentication or Workload Identity Federation available for programmatic access.
使用场景
“Conversational application features”
Build chat and other multi-turn interactions by sending messages to Claude and extending the integration with system prompts, streaming, and client SDK support.
“Large-volume asynchronous processing”
Submit substantial numbers of message requests through the Message Batches API when results do not need to be returned synchronously.
“Stateful agent workflows”
Create managed agent sessions that maintain state in cloud sandboxes, using reusable agent configurations and environment templates for agent-based applications.
“File-based application workflows”
Upload files once and use them across API calls, supporting applications that need Claude to work with shared file inputs.
“Developer prototyping and evaluation”
Use the Console playground, starter documentation, and Cookbook examples to make an initial API call, test prompts, and explore tools, multimodal capabilities, and context-management patterns.
常见问题
What is required to make a first API call?
You need a Claude Console account and an API key. The documented quickstart also requires a development environment and an Anthropic client SDK or direct API request.
Which APIs are available on Claude Platform?
The documented APIs include Messages, Message Batches, Token Counting, Models, Files, and Skills. Agents, Sessions, and Environments are listed as beta APIs for managed agent workflows.
How can applications authenticate?
Requests can use an API key or a short-lived access token obtained through Workload Identity Federation. The API uses standard authentication and version headers; official SDKs manage several of these headers automatically.
Can Claude Platform be used through a cloud provider?
Yes. Claude is available through AWS, Google Cloud, and Microsoft Azure, in addition to the direct Claude API. Feature availability, authentication, billing, and support vary by platform.
Does the source specify pricing?
The provided pricing page does not contain pricing details, so no prices, plan limits, or free-tier claims can be confirmed from the available source.
快速信息
- Category
- Developer platform and AI API
- Primary users
- Developers and teams building applications and agents
- API style
- RESTful API
- Direct API endpoint
- api.anthropic.com
- Core access
- Claude models and Claude Managed Agents
- Setup
- Claude Console account plus API key or Workload Identity Federation
Claude Platform 替代品
Pipecat
www.pipecat.ai
Pipecat is an open-source Python framework and ecosystem for building real-time voice and multimodal AI agents. It helps developers compose pipelines that coordinate speech, language, vision, video, text, tools, and telephony transports.
Phonic
phonic.ai
Speech-to-speech voice agent infrastructure for teams building phone and audio workflows.
NetMind
netmind.ai
Unified AI APIs, GPU compute, agent platforms, and managed solutions for teams building and running AI systems.
AgentOps
www.agentops.ai
AgentOps is a developer platform for building, tracing, debugging, and deploying AI agents and LLM applications. It provides observability for agents built with OpenAI, CrewAI, AutoGen, and more than 400 LLMs and frameworks.
Graphiti
www.getzep.com
Graphiti is an open-source GitHub project for building real-time knowledge graphs for AI agents. It provides a codebase that developers can inspect, extend, and contribute to through the repository’s development workflow.
Agent Router
theagentrouter.ai
Agent Router is an open-source Go project for managing unified access to generative AI services through Envoy Gateway. It is intended for teams that want a gateway-based control point for AI service traffic and related agent or MCP workloads.