Skip to main content

Sarvam 30B (High)

SarvamOpen WeightApache 2.0 · Commercial OK

Description

Sarvam-30B is an open-source 30B-parameter Mixture-of-Experts reasoning model from Sarvam AI trained from scratch and optimized for Indian languages, coding, and conversational workloads. It uses 128 sparse experts with 2.4B active parameters per token, Grouped Query Attention, and was pre-trained on 16 trillion tokens spanning code, mathematics, multilingual, and web data.

Release Date
2026-03-06
Parameters
30.0B
Context Length
66K
Modalities
text

Capability Radar

70
general
60
coding
80
reasoning
60
scienceest.
40
agents
0
multimodal

Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.

Rankings

No ranking data available

Benchmark Scores (LLM Stats)

(LLM Stats (zeroeval))

General

MMLU85.1%SR
Arena-Hard v249.0%SR

Language

MMLU-Pro80.0%SR

Math

MATH-50097.0%SR
AIME 202596.7%SR
HMMT2574.2%SR
HMMT 202573.3%SR
Beyond AIME58.3%SR

Reasoning

MBPP0.93 / 100SR
HumanEvalOpenAI (2021)92.1%SR
LiveCodeBench v670.0%SR
GPQANYU + Cohere + Anthropic (2023)66.5%SR
BrowseCompOpenAI (2025)35.5%SR
SWE-Bench Verified34.0%SR

AA Evaluation Indices

(Artificial Analysis)

No AA evaluation data available

LLM Stats Category Scores

(LLM Stats (zeroeval))
Language
80
Legal
80
Math
80
Finance
80
Healthcare
80
Physics
70
Reasoning
70
General
70
Biology
70
Chemistry
70
Code
60
Creativity
50
Writing
50
Search
40
Agents
40
Frontend Development
30

Pricing

No pricing data available

Speed

No speed data available

Provider Price Ranking

No provider data available

External Sources