Developer docs
ModelOne serves text chat for an allowlisted set of OpenAI and Anthropic models. The base URL is https://modelone.app/api/v1. ModelOne is not affiliated with or endorsed by OpenAI, Anthropic, or OpenRouter.
What is implemented
- Sign-in, API key creation, and key revocation.
GET /api/v1/modelswith a ModelOne API key.POST /api/v1/chat/completionsfor JSON and streaming text. A call is refused until inference is enabled on the server.- The signed-in Playground and usage history, under the same gate.
What stays off
A completion spends credits only when MODELONE_INFERENCE_ENABLED=true. Production also requires MODELONE_INFERENCE_PRODUCTION_APPROVED=true. Until then, the chat route returns 403 with code disabled. The model list still responds and includes inference_enabled.
Quickstart
Create an account, store a key, and send a first text request.
API reference
Authentication, request fields, streaming, errors, and limits.
Models
The server allowlist, with ids, text capability, and published rates.
Agent setup
How a person can approve setup in Cursor, Claude Code, Codex, or another agent.
Skip to main content