When was Sarvam-105B released?

Sarvam-105B was released on March 6, 2026 by sarvamai. This is the official Sarvam-105B release date tracked on LLM Stats.

Is Sarvam-105B available via API?

Yes, Sarvam-105B is available via API. See the official documentation for authentication and endpoint details.

How big is Sarvam-105B?

Sarvam-105B has 105 billion parameters. It was trained on 12.0 trillion tokens. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Sarvam-105B?

Sarvam-105B was created by sarvamai.

What is the license for Sarvam-105B?

Sarvam-105B is released under the Apache 2.0 license. This is an open-source / open-weight license that permits self-hosting.

Where is the Sarvam-105B paper or technical report?

Sarvam-105B has a paper or technical report available at https://www.sarvam.ai/blogs/sarvam-30b-105b. Use that source for architecture, training, release and evaluation details.

What models should I compare Sarvam-105B against?

Common Sarvam-105B comparisons include Sarvam-105B vs GPT-5.1 Codex High, Sarvam-105B vs o1-pro, Sarvam-105B vs Nova 2 Pro. Compare them side by side for benchmark scores, pricing, context window, latency and API availability.

Sarvam-105B API Pricing, Context Window & Benchmarks

Name: Sarvam-105B
Author: sarvamai

Sarvam-105B: API Pricing, Context Window & Benchmarks

Sarvam-105B is a language model from Unknown Organization, released in March 2026.

Sarvam-105B is Sarvam AI's flagship open-source Mixture-of-Experts reasoning model built for complex reasoning, coding, and agentic workflows. It uses 128 sparse experts with Multi-head Latent Attention for efficient long-context inference

Sarvam-105B benchmarks

Rankings

Quality Tracker

Sarvam-105B Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

llm-stats.com - Wed Jul 29 2026

Notice missing or incorrect data?

Sarvam-105B model size

Sarvam-105B has 105 billion parameters and was trained on 12 trillion tokens. See how it compares to other models in the same parameter range.

ParametersTraining tokens

105BMoE

12Ttokens

114× tokens-to-params ratio

Very large (80–200B)

105B

1B7B70B405B

Sarvam-105B API

Available from the model provider

Sarvam-105B has an official provider API. It is not currently routed through the LLM Stats gateway.

Read the official API documentation

Sarvam-105B latency

Sarvam-105B time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Sarvam-105B examples

Recent arena outputs from Sarvam-105B, picked from the highest-ranked matchups.

Sarvam-105B license

Sarvam-105B is released under the Apache 2.0 license, which permits commercial use, has 105.0B parameters.

License: Apache 2.0; Commercial use allowed
Parameters: 105.0B

Apache License 2.0 - allows commercial use

Sarvam-105B resources

Official sources for Sarvam-105B: api documentation, official playground, official launch post, model weights.

Sarvam-105B vs other models

The most-compared alternatives to Sarvam-105B are GPT-5.1 Codex High, o1-pro, Nova 2 Pro. Open any pair side-by-side for benchmarks, pricing, context, and latency.