Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
DeepSeekMath V2
63
Winner · 2/8 categoriesSarvam 30B
48
2/8 categoriesDeepSeekMath V2· Sarvam 30B
Pick DeepSeekMath V2 if you want the stronger benchmark profile. Sarvam 30B only becomes the better choice if knowledge is the priority.
DeepSeekMath V2 is clearly ahead on the aggregate, 63 to 48. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
DeepSeekMath V2's sharpest advantage is in agentic, where it averages 63.9 against 35.5. The single biggest benchmark swing on the page is BrowseComp, 66% to 35.5%. Sarvam 30B does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.
DeepSeekMath V2 gives you the larger context window at 128K, compared with 64K for Sarvam 30B.
BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.
Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.
| Benchmark | DeepSeekMath V2 | Sarvam 30B |
|---|---|---|
| AgenticDeepSeekMath V2 wins | ||
| Terminal-Bench 2.0 | 65% | — |
| BrowseComp | 66% | 35.5% |
| OSWorld-Verified | 61% | — |
| CodingDeepSeekMath V2 wins | ||
| HumanEval | 72% | 92.1% |
| SWE-bench Verified | 45% | 34% |
| LiveCodeBench | 44% | — |
| SWE-bench Pro | 51% | — |
| LiveCodeBench v6 | — | 70.0% |
| Multimodal & Grounded | ||
| MMMU-Pro | 64% | — |
| OfficeQA Pro | 73% | — |
| Reasoning | ||
| MuSR | 75% | — |
| BBH | 86% | — |
| LongBench v2 | 75% | — |
| MRCRv2 | 72% | — |
| gpqaDiamond | — | 66.5% |
| KnowledgeSarvam 30B wins | ||
| GPQA | 79% | — |
| SuperGPQA | 77% | — |
| FrontierScience | 73% | — |
| SimpleQA | 77% | — |
| MMLU | — | 85.1% |
| MMLU-Pro | — | 80% |
| Instruction Following | ||
| IFEval | 83% | — |
| Multilingual | ||
| MGSM | 87% | — |
| MMLU-ProX | 80% | — |
| MathematicsSarvam 30B wins | ||
| AIME 2024 | 82% | — |
| AIME 2025 | 81% | 80% |
| HMMT Feb 2023 | 76% | — |
| HMMT Feb 2024 | 78% | — |
| HMMT Feb 2025 | 77% | — |
| BRUMO 2025 | 79% | — |
| MATH-500 | 90% | 97% |
| HMMT Feb 2025 | — | 73.3% |
| HMMT Nov 2025 | — | 74.2% |
DeepSeekMath V2 is ahead overall, 63 to 48. The biggest single separator in this matchup is BrowseComp, where the scores are 66% and 35.5%.
Sarvam 30B has the edge for knowledge tasks in this comparison, averaging 80 versus 76.1. DeepSeekMath V2 stays close enough that the answer can still flip depending on your workload.
DeepSeekMath V2 has the edge for coding in this comparison, averaging 46.9 versus 34. Inside this category, HumanEval is the benchmark that creates the most daylight between them.
Sarvam 30B has the edge for math in this comparison, averaging 86.5 versus 82.6. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
DeepSeekMath V2 has the edge for agentic tasks in this comparison, averaging 63.9 versus 35.5. Inside this category, BrowseComp is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.