Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Granite-4.0-350M
~27
0/8 categoriesMistral 8x7B v0.2
28
Winner · 3/8 categoriesGranite-4.0-350M· Mistral 8x7B v0.2
Pick Mistral 8x7B v0.2 if you want the stronger benchmark profile. Granite-4.0-350M only becomes the better choice if its workflow or ecosystem matters more than the raw scoreboard.
Mistral 8x7B v0.2 finishes one point ahead overall, 28 to 27. That is enough to call, but not enough to treat as a blowout. This matchup comes down to a few meaningful edges rather than one model dominating the board.
Mistral 8x7B v0.2's sharpest advantage is in multilingual, where it averages 58.8 against 16.2. The single biggest benchmark swing on the page is MGSM, 16.2% to 62%.
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 | Granite-4.0-350M | Mistral 8x7B v0.2 |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 24% |
| BrowseComp | — | 34% |
| OSWorld-Verified | — | 28% |
| Coding | ||
| HumanEval | 38% | 21% |
| SWE-bench Verified | — | 16% |
| LiveCodeBench | — | 12% |
| SWE-bench Pro | — | 14% |
| Multimodal & Grounded | ||
| MMMU-Pro | — | 26% |
| OfficeQA Pro | — | 40% |
| Reasoning | ||
| BBH | 33.3% | 62% |
| MuSR | — | 25% |
| LongBench v2 | — | 39% |
| MRCRv2 | — | 38% |
| KnowledgeMistral 8x7B v0.2 wins | ||
| MMLU | 36.2% | 29% |
| GPQA | 26.1% | 28% |
| MMLU-Pro | 14.4% | 52% |
| SuperGPQA | — | 26% |
| HLE | — | 3% |
| FrontierScience | — | 35% |
| SimpleQA | — | 27% |
| Instruction FollowingMistral 8x7B v0.2 wins | ||
| IFEval | 61.6% | 67% |
| MultilingualMistral 8x7B v0.2 wins | ||
| MGSM | 16.2% | 62% |
| MMLU-ProX | — | 57% |
| Mathematics | ||
| AIME 2023 | — | 29% |
| AIME 2024 | — | 31% |
| AIME 2025 | — | 30% |
| HMMT Feb 2023 | — | 25% |
| HMMT Feb 2024 | — | 27% |
| HMMT Feb 2025 | — | 26% |
| BRUMO 2025 | — | 28% |
| MATH-500 | — | 59% |
Mistral 8x7B v0.2 is ahead overall, 28 to 27. The biggest single separator in this matchup is MGSM, where the scores are 16.2% and 62%.
Mistral 8x7B v0.2 has the edge for knowledge tasks in this comparison, averaging 28.4 versus 18.5. Inside this category, MMLU-Pro is the benchmark that creates the most daylight between them.
Mistral 8x7B v0.2 has the edge for instruction following in this comparison, averaging 67 versus 61.6. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Mistral 8x7B v0.2 has the edge for multilingual tasks in this comparison, averaging 58.8 versus 16.2. Inside this category, MGSM 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.