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
7
general
19
coding
63
reasoning
37
scienceest.
40
agents
0
multimodal
Science uses a reasoning proxy when dedicated science benchmarks are unavailable.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Agentic Capability | 109 | 36.0 | LS |
| Code Ranking | 491 | 2.0 | AA |
| General Ranking | 457 | 19.0 | AA |
| Science | 321 | 35.0 | AA |
Benchmark Scores (LLM Stats)
Agents
BrowseComp
35.5%SR
Biology
GPQA
66.5%SR
Code
HumanEval
92.1%SR
SWE-Bench Verified
34.0%SR
Creativity
Arena-Hard v2
49.0%SR
Finance
MMLU
85.1%SR
MMLU-Pro
80.0%SR
General
MBPP
0.93 / 100SR
LiveCodeBench v6
70.0%SR
Math
MATH-500
97.0%SR
AIME 2025
96.7%SR
HMMT25
74.2%SR
HMMT 2025
73.3%SR
Beyond AIME
58.3%SR
AA Evaluation Indices
Intelligence Index6.6
Gpqa0.6
Tau20.3
Ifbench0.3
Scicode0.2
Hle0.1
Terminalbench Hard0.0
Lcr0.0
LLM Stats Category Scores
Language80
Legal80
Math80
Finance80
Healthcare80
Physics70
Reasoning70
General70
Biology70
Chemistry70
Code60
Creativity50
Writing50
Search40
Agents40
Frontend Development30
Pricing
Input Price$0.026 / 1M tokens
Output Price$0.11 / 1M tokens
Blended Price (3:1)$0.047 / 1M tokens
Speed
Tokens/sec188.0
Time to First Token1.14s
Time to Answer11.78s
Provider Price Ranking
Provider Price Ranking
3 providers
Cheapest: FastRouterMost Expensive: NanoGPT
ProviderInputOutput
1FastRouterCheapest
$0.02
$0.1
2SarvamPRIMARY
$0.026
$0.11
3NanoGPT
$0.028
$0.111
Compare pricing across different API providers for this model.