Repository navigation
[FEATURE]: Hydrate custom provider model metadata from /v1/models and Models.dev #25407
Description
Feature hasn't been suggested before.
- I have verified this feature I'm about to request hasn't been suggested before.
Describe the enhancement you want to request
Please improve custom OpenAI-compatible providers so model capabilities do not have to be entered manually for every model.
My custom provider exposes models through /v1/models, but OpenCode still needs each model manually configured in opencode.jsonc to get useful capabilities such as:
limit.contextlimit.outputmodalities.input, especially image support- reasoning variants/options for known OpenAI GPT/Codex models
The practical problem: I had an image-capable GPT model exposed by a custom provider, but OpenCode treated it as text-only until I manually added:
"modalities": {
"input": ["text", "image"],
"output": ["text"]
}The context display also showed unknown capacity until I manually added limits.
For a provider with many models, manually adding all model metadata is very tedious and easy to get wrong.
Suggested behavior:
- For custom OpenAI-compatible providers, call
{baseURL}/modelsto discover available model IDs. - If a model ID is known in Models.dev, hydrate OpenCode metadata from there where possible.
- Merge this with user config, with explicit user config taking precedence.
- At minimum, auto-fill known limits and modalities for recognized models.
Related issues that seem to describe the same general pain point:
- Auto-discover models from OpenAI-compatible provider endpoints #6231
- [Feature Request]: Allow dynamic model passthrough for custom OpenAI-compatible providers without explicit declaration #18219
- [FEATURE]:LM Studio provider should auto-detect models via /v1/models API (similar to Continue extension) #23327
This issue is specifically about avoiding manual capability metadata for custom OpenAI-compatible providers, because missing metadata can make supported features, such as image input, appear broken.
Activity
github-actions commentedon May 2, 2026
This issue doesn't fully meet our contributing guidelines.
What needs to be fixed:
[FEATURE]:and include the required verification checkbox and description fields.Please edit this issue to address the above within 2 hours, or it will be automatically closed.
This issue might be a duplicate of existing issues. Please check:
If you believe this was flagged incorrectly, please let a maintainer know.
[-]Custom providers should auto-discover model metadata from /v1/models[/-][+][FEATURE]: Hydrate custom provider model metadata from /v1/models and Models.dev[/+]github-actions commentedon May 2, 2026
Thanks for updating the issue! It now follows the contributing guidelines and feature request template. The
needs:compliancelabel has been removed.github-actions commentedon Jul 3, 2026
To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.