Sarvam AI
Sarvam chat, streaming text-to-speech and speech-to-text models.
Route Sarvam AI models through DeepintShield. The gateway sends your Sarvam subscription key in both authentication forms Sarvam expects.
What works
- List models and Chat Completions with SSE streaming
- Text-to-speech with streaming audio chunks
- Speech-to-text (non-streaming)
- Tools, structured output and reasoning effort
Limits to know: Responses, embeddings, files, batches, image operations and Realtime are rejected locally as unsupported.
How to connect
Add your Sarvam API subscription key in Model Hub → AI Providers, create a Virtual Key, then call sarvam/<model>.
Read the Sarvam AI provider docs. Requests get the same Virtual Key budgets, guardrails and logs as every other provider on the AI gateway.