Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Llama 4 Behemoth
34
0/8 categoriesSarvam 30B
48
Winner · 4/8 categoriesLlama 4 Behemoth· Sarvam 30B
Pick Sarvam 30B if you want the stronger benchmark profile. Llama 4 Behemoth only becomes the better choice if you would rather avoid the extra latency and token burn of a reasoning model.
Sarvam 30B is clearly ahead on the aggregate, 48 to 34. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Sarvam 30B's sharpest advantage is in knowledge, where it averages 80 against 37.3. The single biggest benchmark swing on the page is HumanEval, 40% to 92.1%.
Sarvam 30B is the reasoning model in the pair, while Llama 4 Behemoth is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Sarvam 30B gives you the larger context window at 64K, compared with 32K for Llama 4 Behemoth.
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 | Llama 4 Behemoth | Sarvam 30B |
|---|---|---|
| AgenticSarvam 30B wins | ||
| Terminal-Bench 2.0 | 33% | — |
| BrowseComp | — | 35.5% |
| CodingSarvam 30B wins | ||
| HumanEval | 40% | 92.1% |
| SWE-bench Verified | 15% | 34% |
| LiveCodeBench | 13% | — |
| SWE-bench Pro | 15% | — |
| LiveCodeBench v6 | — | 70.0% |
| Multimodal & Grounded | ||
| OfficeQA Pro | 49% | — |
| Reasoning | ||
| MuSR | 44% | — |
| BBH | 62% | — |
| LongBench v2 | 46% | — |
| MRCRv2 | 46% | — |
| gpqaDiamond | — | 66.5% |
| KnowledgeSarvam 30B wins | ||
| MMLU | 48% | 85.1% |
| GPQA | 47% | — |
| SuperGPQA | 45% | — |
| MMLU-Pro | 54% | 80% |
| HLE | 3% | — |
| FrontierScience | 43% | — |
| SimpleQA | 46% | — |
| Instruction Following | ||
| IFEval | 68% | — |
| Multilingual | ||
| MMLU-ProX | 61% | — |
| MathematicsSarvam 30B wins | ||
| AIME 2023 | 48% | — |
| AIME 2024 | 50% | — |
| HMMT Feb 2024 | 46% | — |
| HMMT Feb 2025 | 45% | — |
| BRUMO 2025 | 47% | — |
| MATH-500 | 60% | 97% |
| AIME 2025 | — | 80% |
| HMMT Feb 2025 | — | 73.3% |
| HMMT Nov 2025 | — | 74.2% |
Sarvam 30B is ahead overall, 48 to 34. The biggest single separator in this matchup is HumanEval, where the scores are 40% and 92.1%.
Sarvam 30B has the edge for knowledge tasks in this comparison, averaging 80 versus 37.3. Inside this category, MMLU is the benchmark that creates the most daylight between them.
Sarvam 30B has the edge for coding in this comparison, averaging 34 versus 14.2. 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 52.4. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Sarvam 30B has the edge for agentic tasks in this comparison, averaging 35.5 versus 33. Llama 4 Behemoth stays close enough that the answer can still flip depending on your workload.
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.