Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Nemotron 3 Ultra 500B
60
1/8 categoriesSarvam 105B
60
4/8 categoriesNemotron 3 Ultra 500B· Sarvam 105B
Treat this as a split decision. Nemotron 3 Ultra 500B makes more sense if agentic is the priority or you need the larger 10M context window; Sarvam 105B is the better fit if knowledge is the priority.
Nemotron 3 Ultra 500B and Sarvam 105B 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.
Nemotron 3 Ultra 500B gives you the larger context window at 10M, compared with 128K for Sarvam 105B.
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 | Nemotron 3 Ultra 500B | Sarvam 105B |
|---|---|---|
| AgenticNemotron 3 Ultra 500B wins | ||
| Terminal-Bench 2.0 | 63% | — |
| BrowseComp | 69% | 49.5% |
| OSWorld-Verified | 58% | — |
| CodingSarvam 105B wins | ||
| HumanEval | 66% | — |
| SWE-bench Verified | 42% | 45% |
| LiveCodeBench | 41% | — |
| SWE-bench Pro | 47% | — |
| LiveCodeBench v6 | — | 71.7% |
| Multimodal & Grounded | ||
| MMMU-Pro | 61% | — |
| OfficeQA Pro | 74% | — |
| Reasoning | ||
| MuSR | 69% | — |
| BBH | 85% | — |
| LongBench v2 | 81% | — |
| MRCRv2 | 85% | — |
| gpqaDiamond | — | 78.7% |
| hle | — | 11.2% |
| KnowledgeSarvam 105B wins | ||
| GPQA | 73% | — |
| SuperGPQA | 71% | — |
| MMLU-Pro | 73% | 81.7% |
| HLE | 15% | — |
| FrontierScience | 67% | — |
| MMLU | — | 90.6% |
| Instruction FollowingSarvam 105B wins | ||
| IFEval | 84% | 84.8% |
| Multilingual | ||
| MGSM | 81% | — |
| MMLU-ProX | 79% | — |
| MathematicsSarvam 105B wins | ||
| AIME 2023 | 74% | — |
| HMMT Feb 2023 | 70% | — |
| HMMT Feb 2024 | 72% | — |
| MATH-500 | 84% | 98.6% |
| AIME 2025 | — | 88.3% |
| HMMT Feb 2025 | — | 85.8% |
| HMMT Nov 2025 | — | 85.8% |
Nemotron 3 Ultra 500B and Sarvam 105B are tied on overall score, so the right pick depends on which category matters most for your use case.
Sarvam 105B has the edge for knowledge tasks in this comparison, averaging 81.7 versus 56.1. Inside this category, MMLU-Pro is the benchmark that creates the most daylight between them.
Sarvam 105B has the edge for coding in this comparison, averaging 45 versus 43.5. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.
Sarvam 105B has the edge for math in this comparison, averaging 92.3 versus 84. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Nemotron 3 Ultra 500B has the edge for agentic tasks in this comparison, averaging 62.8 versus 49.5. Inside this category, BrowseComp is the benchmark that creates the most daylight between them.
Sarvam 105B has the edge for instruction following in this comparison, averaging 84.8 versus 84. Inside this category, IFEval 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.