Repository navigation
[Feature]: Add Cheaper Inference as a named compatible provider #501
Description
Summary
Add Cheaper Inference as a named compatible provider preset, like OrcaRouter (#387).
Problem / goal
Cheaper Inference is an OpenAI-compatible LLM gateway.
One API key gives access to models from several labs.
Today, users must add it as a custom provider with a base URL and model ids.
Primary user(s)
- Bob (IT / power user)
- Susan (non-technical)
- Other team roles
OpenCode primitive alignment
OpenCode supports OpenAI-compatible providers through @ai-sdk/openai-compatible.
The preset only fills the provider config: base URL https://api.cheaperinference.com/v1 and a model list.
Alignment with VISION/PRINCIPLES/PRODUCT
It uses the existing compatible provider preset path. It adds no new engine or dependency.
Testability
Unit tests for the preset and a store-level connection test, same as #387.
Manual: open Providers, choose Cheaper Inference, paste a key, and send a prompt with gpt-5.4-mini.
Ready to build it yourself?
Yes
Activity