Available at modelplane.dev. Manage multiple providers, model groups, routing policies, coding plans, thinking output, and usage billing through one OpenAI-compatible entry point.
This documentation is written for end users of One AI Gateway. It focuses on how to use the console, how to organize model groups and routing, how usage is tracked, and what developers need for LLM calls.
Hide provider-specific credentials, hosts, and model names behind a stable model group.
Send free, paid, and internal users to different backends using request metadata.
Combine coding-plan backends with fallback targets so quota issues do not interrupt users.
Use one reasoning request format and read standardized reasoning text from responses when available.
Start with the Quickstart, then create your first Model Group.