I’m building Your Model, a third-party OpenAI-compatible multi-model API for developers who want one API key and one prepaid balance across supported routes.
I built it because “the request returned 200” is not enough for an agent workflow. The useful checks are:
- Chat Completions and Responses behavior
- ordered streaming deltas
- tool-call state
- visible usage and billing records
- model IDs and error behavior
The current public product includes:
- OpenAI-compatible Chat Completions and Responses endpoints
- streaming and tool-call support
- public pricing and usage logs
- 17 currently listed public model routes
I also ran a scoped Codex workflow with Codex CLI 0.132.0, gpt-5.4-mini, and the Responses API. Basic response, streaming, file read, and a minimal file edit passed. Independent function-tool continuation has not been separately verified.
Current limitations:
- support depends on the model route and provider path
- OpenAI-compatible does not mean every agent feature behaves identically
- this is a hosted third-party service, not a self-hosted gateway
I’m sharing this as a project showcase, not as a claim that every client/model combination is compatible. Feedback on the compatibility checks developers actually need would be useful.
Website: https://y-models.com/?utm_source=dev&utm_medium=community&utm_campaign=distribution&utm_content=showdev
Pricing: https://y-models.com/pricing
Docs: https://docs.y-models.com
Disclosure: I’m building Your Model.
Top comments (0)