Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Kimi K2
53
3/8 categoriesSarvam 105B
60
Winner · 2/8 categoriesKimi K2· Sarvam 105B
Pick Sarvam 105B if you want the stronger benchmark profile. Kimi K2 only becomes the better choice if coding is the priority or you would rather avoid the extra latency and token burn of a reasoning model.
Sarvam 105B is clearly ahead on the aggregate, 60 to 53. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Sarvam 105B's sharpest advantage is in mathematics, where it averages 92.3 against 67.9. The single biggest benchmark swing on the page is AIME 2025, 49.5% to 88.3%. Kimi K2 does hit back in coding, so the answer changes if that is the part of the workload you care about most.
Sarvam 105B is the reasoning model in the pair, while Kimi K2 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.
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 | Kimi K2 | Sarvam 105B |
|---|---|---|
| AgenticKimi K2 wins | ||
| Terminal-Bench 2.0 | 47.1% | — |
| BrowseComp | 60.2% | 49.5% |
| Tau2-Telecom | 66.1% | — |
| CodingKimi K2 wins | ||
| SWE-bench Verified | 65.8% | 45% |
| LiveCodeBench | 53.7% | — |
| LiveCodeBench v6 | — | 71.7% |
| Multimodal & Grounded | ||
| Coming soon | ||
| Reasoning | ||
| hle | 44.9% | 11.2% |
| gpqaDiamond | — | 78.7% |
| KnowledgeSarvam 105B wins | ||
| MMLU | 89.5% | 90.6% |
| GPQA | 75.1% | — |
| SuperGPQA | 57.2% | — |
| MMLU-Pro | 81.1% | 81.7% |
| SimpleQA | 31% | — |
| Instruction FollowingKimi K2 wins | ||
| IFEval | 89.8% | 84.8% |
| Multilingual | ||
| sweMultilingual | 61.1% | — |
| MathematicsSarvam 105B wins | ||
| AIME 2024 | 69.6% | — |
| AIME 2025 | 49.5% | 88.3% |
| MATH-500 | 97.4% | 98.6% |
| HMMT Feb 2025 | 38.8% | — |
| HMMT Feb 2025 | — | 85.8% |
| HMMT Nov 2025 | — | 85.8% |
Sarvam 105B is ahead overall, 60 to 53. The biggest single separator in this matchup is AIME 2025, where the scores are 49.5% and 88.3%.
Sarvam 105B has the edge for knowledge tasks in this comparison, averaging 81.7 versus 64. Inside this category, MMLU is the benchmark that creates the most daylight between them.
Kimi K2 has the edge for coding in this comparison, averaging 58.2 versus 45. 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 67.9. Inside this category, AIME 2025 is the benchmark that creates the most daylight between them.
Kimi K2 has the edge for agentic tasks in this comparison, averaging 52.1 versus 49.5. Inside this category, BrowseComp is the benchmark that creates the most daylight between them.
Kimi K2 has the edge for instruction following in this comparison, averaging 89.8 versus 84.8. 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.