Sarvam AI model product
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.
Updated Aug 10, 2026. Default version: Sarvam-30B
Structured fields from the published default version.
This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.
Published benchmark records for the scored version Sarvam-30B.
| MBPP | 0.9 | 1 | 37 | 100.0% | C | |
| Beyond AIME | 0.6 | 5 | 5 | 0.0% | C | |
| MATH-500 | 1.0 | 9 | 32 | 74.2% | C | |
| HumanEval | 0.9 | 10 | 76 | 88.0% | C | |
| Arena-Hard v2 | 0.5 | 15 | 16 | 6.7% | C | |
| HMMT25 | 0.7 | 15 | 25 | 41.7% | C | |
| AIME 2025 | 1.0 | 18 | 114 | 85.0% | C | |
| HMMT 2025 | 0.7 | 28 | 33 | 15.6% | C | |
| LiveCodeBench v6 | 0.7 | 30 | 53 | 44.2% | C | |
| MMLU | 0.9 | 43 | 100 | 57.6% | C | |
| BrowseComp | 0.4 | 53 | 58 | 8.8% | C | |
| MMLU-Pro | 0.8 | 55 | 129 | 57.8% | C | |
| SWE-Bench Verified | 0.3 | 98 | 104 | 5.8% | C | |
| GPQA | 0.7 | 142 | 233 | 39.2% | C |
Preference and agent-evaluation signals from published Arena datasets.
The default version has no published Arena rows, or its source alias has not been resolved.
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| NanoGPT | sarvam-30b | global | $0.028 | $0.111 | 65.5K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.
| Sarvam-30B | 47.0 | 30B | 128K | 128K | Yes | Apache 2.0 |
A concise description based on the published model registry.
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.
Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.
Data snapshot: 2026-08-07. Editorial model content is not available in the backend.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest published LLMBoard score.
Common questions about Sarvam 30B.
Sarvam 30B's default version was released on Mar 6, 2026.
No official standard PAYG price is currently available for Sarvam 30B. The lowest tracked third-party offer starts at $0.028 input and $0.111 output via NanoGPT.
Sarvam 30B is published under Sarvam AI in the model registry.
The default version has a 128K token context window.
Yes. The default version is marked as open weight under Apache 2.0.
1 published provider offerings are linked to the default version.
Nearby ranked alternatives include Gemma 4 12B, Phi 4 Reasoning, Gemma 2 27B.