LLM Inference
Rlab Relay
First-party LLM inference through Rlab Relay with multiple compatible API styles.
- Live smoke
- Tiny Responses, Messages, Chat Completions, Gemini generateContent, and streamed chat smoke paths are declared for Rlab Relay.
- Adapter coverage
- Native first-party model adapter is registered for Responses, Messages, Chat Completions, Gemini generateContent, and SSE streaming.
- Operation depth
- Full Rlab compatibility route matrix is implemented with mocked contracts, token usage normalization, gateway routes, and tiny live-smoke coverage.
- Next deep test
- Add model catalog pricing resolution and richer managed-provider cost metadata.
- Stage
- Operate
- Risk
- write
- Governance
- review required
- approval required
- approval optional
- Credential
- Unknown
- Last smoke
- Unknown
- Local env
- Missing
- Database
- None
Credential
Connected without secret displayOperations
Enablement and default cost1gemini_generate_content3responsesmessageschat_completions