# Prism Inference Documentation

> OpenAI Chat Completions and Anthropic Messages compatible inference at https://api.prisminference.com.

## API summary

- OpenAI base URL: https://api.prisminference.com/v1
- Anthropic base URL: https://api.prisminference.com
- POST /v1/chat/completions
- GET /v1/models
- GET /v1/models/{model}
- POST /v1/messages
- POST /v1/messages/count_tokens
- Authentication: Authorization: Bearer PRISM_API_KEY or x-api-key: PRISM_API_KEY

## Get started

- [Introduction](https://docs.prisminference.com/introduction): Supported formats, base URLs, and the Prism inference surface.
- [Quickstart](https://docs.prisminference.com/quickstart): Send a request with cURL, TypeScript, Python, or the Anthropic SDK.
- [Models](https://docs.prisminference.com/models): Public model IDs and context windows.
- [Authentication](https://docs.prisminference.com/authentication): API key headers, base URLs, and key handling.

## API reference

- [Chat Completions](https://docs.prisminference.com/api-reference/chat-completions): OpenAI-compatible requests, responses, and streaming chunks.
- [Anthropic Messages](https://docs.prisminference.com/api-reference/messages): Message creation and input token counting.
- [OpenAPI](https://docs.prisminference.com/api-reference/openapi): Machine-readable OpenAPI 3.1 contract.

## Guides

- [Streaming](https://docs.prisminference.com/guides/streaming): OpenAI and Anthropic streaming, disconnects, and retries.
- [Tool calling](https://docs.prisminference.com/guides/tool-calling): OpenAI function tools and Anthropic tool-use loops.
- [Structured outputs](https://docs.prisminference.com/guides/structured-outputs): JSON mode and strict JSON Schema.
- [Prompt caching](https://docs.prisminference.com/guides/prompt-caching): Stable prompt prefixes and conversation cache keys.
- [Coding agents](https://docs.prisminference.com/guides/coding-agents): Claude Code and OpenAI-compatible harness setup.
- [Errors](https://docs.prisminference.com/errors): Error envelopes, status codes, rate limits, and retry policy.

## Models

- GLM-5.3: prism-glm53, 1M context
- Kimi K3: prism-kimik3, 1M context
- Qwen: prism-qwen, 256K context
- MiniMax: prism-minimax, 1M context
- DeepSeek-V4-Flash: prism-dsv4flash, 1M context

## Machine-readable files

- [OpenAPI 3.1](https://docs.prisminference.com/openapi.yaml)
- [Documentation index](https://docs.prisminference.com/llms.txt)