Developer Docs › Supply lines

Supply linesAlpha

The same model is offered by different supply lines; choose by price, speed and data policy, and pay for the line you use.

Alpha feature: invited accounts only; lines, prices and metrics may change at any time. The "Supply lines" page does not appear in the console until your account is enabled.

Why supply lines

The same model, especially open models such as DeepSeek, Qwen, GLM, MiniMax and Kimi, is often served by several providers with very different price, speed, stability and data policies. A regular group has the platform pick the route for you; supply lines give that choice to you:

  • See before you choose: each line shows the provider type (model vendor / public cloud / third-party host, etc.), processing region, and whether data may be used for training.
  • Measured metrics in the open: time to first token (P95), output speed, 7-day success rate and live status, measured by OneFast API itself.
  • Pay only for the line you pick: each line has its own price (per million tokens), and you are billed for the line that actually served the request.

Lines use anonymous codes and do not map to specific provider names.

How to use

  1. On the "Supply lines" page in the console, compare lines per model and pick one or more.
  2. Click "Create key with selected lines" to create an API key.
  3. Call the API as usual; endpoints and parameters do not change.
  • With several lines, they are used in order: if one is unavailable the next one takes over, and you are billed for the line that served the request.
  • With a single line, only that line is used and requests do not fall back to other lines.