Sarvam-M

Sarvam AIReasoning

sarvam/sarvam-mOpen Weights

Hybrid-reasoning multilingual model for Indian languages built on Mistral Small

Benchmark Pricing$0.00 / 1M in$0.00 / 1M out

Context: 32,768 tokens•Max Output: 32,768 tokens•Released: 2025-05-20

TECHNICAL SPECIFICATIONS

Context Window32,76832.8K tokens

Max Output32,76832.8K tokens

LicenseApache 2.0Self-hostable

Input Modalities

Text

Function / Tool CallingNot Supported

Structured OutputsPrompt Based

MODEL CONTEXT PROTOCOL (MCP) READINESS

⚠ Sarvam-M does not natively advertise tool-calling capabilities. When connecting to MCP servers, function calling must be handled via structured prompt injection.

MCP Context Window Recommendations

With a context capacity of 32,768 tokens, this model can handle compact tool sets. Consider keeping MCP server tool manifests minimal to preserve headroom.

API CALL INVOCATION EXAMPLE

// Example: Calling Sarvam-M via OpenAI SDK
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.SARVAM_API_KEY,
});

const response = await client.chat.completions.create({
  model: "sarvam/sarvam-m",
  messages: [{ role: "user", content: "Hello! Can you inspect MCP tools?" }],
});

console.log(response.choices[0].message.content);