Agent Integrations

Sarvam chat completions (OpenAI-compatible)

Forwards an OpenAI-shaped chat completion to Sarvam. sarvam-105b reasons by default (reasoning_effort: medium) and bills hidden reasoning tokens as output; send reasoning_effort: null to turn it off. Billed at Sarvam's per-token list price (converted at SARVAM_INR_PER_USD), at cost like other LLM inference. The response is Sarvam's unchanged plus an openhuman billing envelope; with stream: true it is an SSE stream ending in a metadata event.

post/agent-integrations/sarvam/chat/completions

Request

  • Base URL: https://api.tinyhumans.ai
  • URL: https://api.tinyhumans.ai/agent-integrations/sarvam/chat/completions
  • Auth: HTTP bearer

Request body

model'sarvam-105b' | 'sarvam-105b-conversations' required
messagesobject[] required
streamboolean
max_tokensinteger
temperaturenumber
reasoning_effort'low' | 'medium' | 'high' nullable
toolsobject[]

Response

OpenAI chat.completion (or SSE chunks) plus the openhuman billing envelope

Changes

No recorded changes to this operation across all 1 revision of this API.