Problem
AI/ML API is OpenAI-compatible and can be configured manually in Zero, but it is not discoverable in the built-in provider catalog or setup wizard. Users must currently enter the endpoint, key environment variable, and model manually.
Requested behavior
Add a focused built-in AI/ML API provider preset using Zero's existing OpenAI-compatible path:
- catalog id / label:
aimlapi / AI/ML API
- position: immediately after the recommended GitLawb OpenGateway entry
- base URL:
https://api.aimlapi.com/v1
- key env:
AIMLAPI_API_KEY
- default model:
openai/gpt-5-chat
- bearer auth and existing OpenAI-compatible chat/tool/streaming runtime
- existing
/models discovery and coding-model filtering
- runtime attribution headers scoped only to this catalog id:
X-AIMLAPI-Partner-ID: GitlawbX-AIMLAPI-Integration-Repo: Gitlawb/zeroX-AIMLAPI-Integration-Version: 1.0.0
The implementation should stay narrow: provider catalog entry, catalog/runtime regression tests, and a short README setup example. No new transport, dependency, or generic provider abstraction.
Why it belongs in Zero
Zero already supports many hosted OpenAI-compatible providers as first-class presets. A named AI/ML API entry gives users one-key access to multiple model families without requiring custom endpoint configuration, while reusing the established provider architecture.
Alternatives or workarounds considered
Users can create a custom-openai-compatible profile today, but that path is manual and does not provide branded setup visibility or stable integration attribution.
Examples, screenshots, or prior discussion
AI/ML API OpenAI SDK compatibility: https://docs.aimlapi.com/quickstart/supported-sdks
API keys: https://aimlapi.com/app/keys
I can implement and test this focused scope. @jatmn, could you confirm the scope and apply issue-approved so the PR complies with CONTRIBUTING.md?
Problem
AI/ML API is OpenAI-compatible and can be configured manually in Zero, but it is not discoverable in the built-in provider catalog or setup wizard. Users must currently enter the endpoint, key environment variable, and model manually.
Requested behavior
Add a focused built-in AI/ML API provider preset using Zero's existing OpenAI-compatible path:
aimlapi/AI/ML APIhttps://api.aimlapi.com/v1AIMLAPI_API_KEYopenai/gpt-5-chat/modelsdiscovery and coding-model filteringX-AIMLAPI-Partner-ID: GitlawbX-AIMLAPI-Integration-Repo: Gitlawb/zeroX-AIMLAPI-Integration-Version: 1.0.0The implementation should stay narrow: provider catalog entry, catalog/runtime regression tests, and a short README setup example. No new transport, dependency, or generic provider abstraction.
Why it belongs in Zero
Zero already supports many hosted OpenAI-compatible providers as first-class presets. A named AI/ML API entry gives users one-key access to multiple model families without requiring custom endpoint configuration, while reusing the established provider architecture.
Alternatives or workarounds considered
Users can create a
custom-openai-compatibleprofile today, but that path is manual and does not provide branded setup visibility or stable integration attribution.Examples, screenshots, or prior discussion
AI/ML API OpenAI SDK compatibility: https://docs.aimlapi.com/quickstart/supported-sdks
API keys: https://aimlapi.com/app/keys
I can implement and test this focused scope. @jatmn, could you confirm the scope and apply
issue-approvedso the PR complies with CONTRIBUTING.md?