# llm402.ai

> Pay-per-request LLM inference over HTTP 402. No accounts, no API keys. OpenAI-compatible. 400+ models across chat, image, and video.

Four payment rails are supported — every 402 response includes all available options:

- **L402** — Bitcoin Lightning. Header: `Authorization: L402 <macaroon>:<preimage>`
- **x402** — USDC on Base (EIP-3009). Header: `Payment-Signature: <base64 payload>`
- **Cashu** — ecash tokens. Header: `X-Cashu: <cashuB token>`. Server enforces a configured mint allowlist (current: https://mint.minibits.cash/bitcoin, https://mint.lnvoltz.com, https://mint.cubabitcoin.org, https://ecashmint.otrta.me, https://mint.mountainlake.io).
- **Prepaid Balance** — bearer tokens. Header: `Authorization: Bearer bal_<token>`

## Docs

- [Full documentation (markdown)](https://llm402.ai/llms-full.txt): all endpoints, payment flows, and examples in one file, suitable for LLM ingestion
- [API reference (HTML)](https://llm402.ai/docs): human-readable version of the same content
- [Model catalog](https://llm402.ai/models): searchable list of all 400+ models with pricing

## Machine-readable endpoints

- [OpenAPI 3.1 spec](https://llm402.ai/.well-known/openapi.json): endpoints, schemas, auth
- [L402 service discovery](https://llm402.ai/.well-known/l402): L402-specific JSON catalog
- [Live model list](https://llm402.ai/v1/models): OpenAI-format JSON, always current
- [Pricing catalog](https://llm402.ai/api/tags): per-model pricing in sats

## Contact

- X / Twitter: https://x.com/llm402ai
- Email: hello [at] llm402 [dot] ai