# Safe AI Labs > An OpenAI and Anthropic compatible API for frontier models. Every request runs in the region you choose, set by the base URL. ## Pricing - [Pricing](https://safeailabs.ai/pricing): API plans with unlimited tokens at a flat monthly price, each capped in tokens per second, and prices per 1M tokens for each model and region. ## Plans - [API Small](https://safeailabs.ai/pricing#plans): Unlimited tokens at up to 25 tokens per second, $79 a month. - [API Medium](https://safeailabs.ai/pricing#plans): Unlimited tokens at up to 100 tokens per second, $299 a month. - [API Large](https://safeailabs.ai/pricing#plans): Unlimited tokens at up to 400 tokens per second, $1,199 a month. - [Free](https://safeailabs.ai/for/chat): The web chat in your account: 30 messages a day, 1 reply at a time, no attachments. Sign-up asks for a card, which is not charged. ## Set up an agent - [OpenClaw](https://safeailabs.ai/for/openclaw): OpenClaw with unlimited tokens at up to 25 tokens per second, from $79 a month. Prompts stay in memory for the session and are never written to disk. - [Hermes Agent](https://safeailabs.ai/for/hermes): Hermes Agent with unlimited tokens at up to 25 tokens per second, from $79 a month. Prompts stay in memory for the session and are never written to disk. - [Claude Code](https://safeailabs.ai/for/claude-code): Claude Code with unlimited tokens at up to 25 tokens per second, from $79 a month. Prompts stay in memory for the session and are never written to disk. - [Paperclip](https://safeailabs.ai/for/paperclip): Paperclip's agents with unlimited tokens at up to 25 tokens per second, from $79 a month. Prompts stay in memory for the session and are never written to disk. - [Sail, a coding agent](https://safeailabs.ai/for/sail): Sail is a coding agent for your terminal. Install it with one command and use it on a monthly plan with unlimited tokens. ## Chat - [Chat](https://safeailabs.ai/for/chat): A web chat on the models we serve, free for 30 messages a day with a card check. The chat in your account stores its conversations until you delete them. ## Docs - [Quickstart](https://safeailabs.ai/docs): Create a key, point any OpenAI client at your region, and send the request you already send. - [Authentication](https://safeailabs.ai/docs/authentication): Send an API key as a bearer token, keep it server-side, and restrict or revoke it from the keys page. - [Regions](https://safeailabs.ai/docs/regions): Pick the region a request runs in by its base URL, and handle a region with no node serving the model. - [Models](https://safeailabs.ai/docs/models): Compare each model's context length and price per region, and list the models a region serves right now. - [Chat completions](https://safeailabs.ai/docs/chat-completions): Send a chat completion in the OpenAI shape, with every request field, an example response and reasoning output. - [Streaming](https://safeailabs.ai/docs/streaming): Stream a response as server-sent events and read the token usage from its final chunk. - [Tools](https://safeailabs.ai/docs/tools): Define functions on a request, read the tool calls the model returns, and send the results back. - [Errors](https://safeailabs.ai/docs/errors): Read the HTTP status, code and message of every error the API returns, and what to do about each. - [Rate limits](https://safeailabs.ai/docs/rate-limits): Size requests against an API plan's tokens per second or the per-key limits per minute, and read the rate limit headers. - [Plans](https://safeailabs.ai/docs/plans): Compare the API plan tiers by tokens per second, see what counts against the cap, and read your usage against it. - [Retention](https://safeailabs.ai/docs/retention): See the whole record a request leaves, and what happens to a prompt after it is processed. - [Compatibility](https://safeailabs.ai/docs/compatibility): Check which OpenAI and Anthropic endpoints the API serves and which it does not. - [Agents and editors](https://safeailabs.ai/docs/integrations): Set up Claude Code, Codex, Cursor, Aider and other agents and editors to send requests to your region. - [Sail](https://safeailabs.ai/docs/sail): Install Sail, sign in, keep it up to date or remove it, and fix a refused request. ## Coding agents - [Claude Code](https://safeailabs.ai/docs/integrations#claude-code): Messages, `/v1/messages`. - [Codex CLI](https://safeailabs.ai/docs/integrations#codex): Responses, `/v1/responses`. - [OpenCode](https://safeailabs.ai/docs/integrations#opencode): Chat completions, `/v1/chat/completions`. - [Pi](https://safeailabs.ai/docs/integrations#pi): Chat completions, `/v1/chat/completions`. - [OpenClaw](https://safeailabs.ai/docs/integrations#openclaw): Chat completions, `/v1/chat/completions`. - [Hermes Agent](https://safeailabs.ai/docs/integrations#hermes): Chat completions, `/v1/chat/completions`. - [Paperclip](https://safeailabs.ai/docs/integrations#paperclip): Chat completions, `/v1/chat/completions`. - [Aider](https://safeailabs.ai/docs/integrations#aider): Chat completions, `/v1/chat/completions`. - [Goose](https://safeailabs.ai/docs/integrations#goose): Chat completions, `/v1/chat/completions`. - [Crush](https://safeailabs.ai/docs/integrations#crush): Chat completions, `/v1/chat/completions`. - [OpenHands](https://safeailabs.ai/docs/integrations#openhands): Chat completions, `/v1/chat/completions`. - [Qwen Code](https://safeailabs.ai/docs/integrations#qwen-code): Chat completions, `/v1/chat/completions`; Responses, `/v1/responses`. - [Warp](https://safeailabs.ai/docs/integrations#warp): Chat completions, `/v1/chat/completions`. ## Editors - [Zed](https://safeailabs.ai/docs/integrations#zed): Chat completions, `/v1/chat/completions`. - [Cursor](https://safeailabs.ai/docs/integrations#cursor): Chat completions, `/v1/chat/completions`. - [Cline, Roo Code, Kilo Code](https://safeailabs.ai/docs/integrations#cline): Chat completions, `/v1/chat/completions`. - [Continue](https://safeailabs.ai/docs/integrations#continue): Chat completions, `/v1/chat/completions`. - [JetBrains AI Assistant](https://safeailabs.ai/docs/integrations#jetbrains): Chat completions, `/v1/chat/completions`. - [GitHub Copilot in VS Code](https://safeailabs.ai/docs/integrations#copilot): Chat completions, `/v1/chat/completions`; Responses, `/v1/responses`; Messages, `/v1/messages`. ## Chat apps - [Open WebUI](https://safeailabs.ai/docs/integrations#open-webui): Chat completions, `/v1/chat/completions`. - [LibreChat](https://safeailabs.ai/docs/integrations#librechat): Chat completions, `/v1/chat/completions`. - [SillyTavern](https://safeailabs.ai/docs/integrations#sillytavern): Chat completions, `/v1/chat/completions`. - [Jan](https://safeailabs.ai/docs/integrations#jan): Chat completions, `/v1/chat/completions`; Messages, `/v1/messages`. - [Msty Studio](https://safeailabs.ai/docs/integrations#msty): Chat completions, `/v1/chat/completions`. ## Frameworks and SDKs - [Vercel AI SDK](https://safeailabs.ai/docs/integrations#ai-sdk): Chat completions, `/v1/chat/completions`; Responses, `/v1/responses`. - [OpenAI Agents SDK](https://safeailabs.ai/docs/integrations#agents-sdk): Responses, `/v1/responses`. - [LangChain](https://safeailabs.ai/docs/integrations#langchain): Chat completions, `/v1/chat/completions`. - [LiteLLM](https://safeailabs.ai/docs/integrations#litellm): Chat completions, `/v1/chat/completions`. - [LlamaIndex](https://safeailabs.ai/docs/integrations#llamaindex): Chat completions, `/v1/chat/completions`. - [CrewAI](https://safeailabs.ai/docs/integrations#crewai): Chat completions, `/v1/chat/completions`. - [Mastra](https://safeailabs.ai/docs/integrations#mastra): Chat completions, `/v1/chat/completions`. - [Pydantic AI](https://safeailabs.ai/docs/integrations#pydantic-ai): Chat completions, `/v1/chat/completions`. - [AutoGen](https://safeailabs.ai/docs/integrations#autogen): Chat completions, `/v1/chat/completions`. - [Semantic Kernel](https://safeailabs.ai/docs/integrations#semantic-kernel): Chat completions, `/v1/chat/completions`. - [Haystack](https://safeailabs.ai/docs/integrations#haystack): Chat completions, `/v1/chat/completions`. ## Workflow tools - [n8n](https://safeailabs.ai/docs/integrations#n8n): Chat completions, `/v1/chat/completions`. - [Dify](https://safeailabs.ai/docs/integrations#dify): Chat completions, `/v1/chat/completions`. - [Flowise](https://safeailabs.ai/docs/integrations#flowise): Chat completions, `/v1/chat/completions`. - [Langflow](https://safeailabs.ai/docs/integrations#langflow): Chat completions, `/v1/chat/completions`. ## Sail - [Sail](https://safeailabs.ai/sail): A coding agent for the terminal, on a monthly plan with unlimited tokens. Install with `curl -fsSL https://safeailabs.ai/install.sh | sh`. - [Sail changelog](https://safeailabs.ai/sail/changelog): The notes for the current release. ## Optional - [Every docs section in one file](https://safeailabs.ai/llms-full.txt)