Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Grok 3 [Beta]
48
1/8 categoriesSarvam 30B
48
2/8 categoriesGrok 3 [Beta]· Sarvam 30B
Treat this as a split decision. Grok 3 [Beta] makes more sense if coding is the priority or you need the larger 128K context window; Sarvam 30B is the better fit if mathematics is the priority or you want the stronger reasoning-first profile.
Grok 3 [Beta] and Sarvam 30B finish on the same overall score, so this is less about a single winner and more about where the edge shows up. The headline says tie; the benchmark table is where the real choice happens.
Sarvam 30B's sharpest advantage is in agentic, where it averages 35.5 against 35.5. The single biggest benchmark swing on the page is HumanEval, 34% to 92.1%. Grok 3 [Beta] does hit back in mathematics, so the answer changes if that is the part of the workload you care about most.
Sarvam 30B is the reasoning model in the pair, while Grok 3 [Beta] 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. Grok 3 [Beta] 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 | Grok 3 [Beta] | Sarvam 30B |
|---|---|---|
| AgenticTie | ||
| Terminal-Bench 2.0 | 32% | — |
| BrowseComp | 41% | 35.5% |
| CodingGrok 3 [Beta] wins | ||
| HumanEval | 34% | 92.1% |
| SWE-bench Verified | 19% | 34% |
| LiveCodeBench | 57% | — |
| LiveCodeBench v6 | — | 70.0% |
| Multimodal & Grounded | ||
| OfficeQA Pro | 47% | — |
| Reasoning | ||
| MuSR | 38% | — |
| BBH | 62% | — |
| LongBench v2 | 53% | — |
| MRCRv2 | 52% | — |
| gpqaDiamond | — | 66.5% |
| KnowledgeSarvam 30B wins | ||
| MMLU | 42% | 85.1% |
| GPQA | 75.4% | — |
| SuperGPQA | 39% | — |
| MMLU-Pro | 79.9% | 80% |
| HLE | 3% | — |
| FrontierScience | 40% | — |
| SimpleQA | 43.6% | — |
| Instruction Following | ||
| IFEval | 67% | — |
| Multilingual | ||
| MGSM | 60% | — |
| MMLU-ProX | 58% | — |
| MathematicsSarvam 30B wins | ||
| AIME 2023 | 42% | — |
| AIME 2024 | 52.2% | — |
| AIME 2025 | 43% | 80% |
| HMMT Feb 2023 | 38% | — |
| HMMT Feb 2025 | 39% | — |
| BRUMO 2025 | 41% | — |
| MATH-500 | 59% | 97% |
| HMMT Feb 2025 | — | 73.3% |
| HMMT Nov 2025 | — | 74.2% |
Grok 3 [Beta] and Sarvam 30B are tied on overall score, so the right pick depends on which category matters most for your use case.
Sarvam 30B has the edge for knowledge tasks in this comparison, averaging 80 versus 44.9. Inside this category, MMLU is the benchmark that creates the most daylight between them.
Grok 3 [Beta] has the edge for coding in this comparison, averaging 42.8 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 46.3. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Grok 3 [Beta] and Sarvam 30B are effectively tied for agentic tasks here, both landing at 35.5 on average.
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.