Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Mistral Small 4 (Reasoning)
~64
0/8 categoriesQwen3.5-27B
71
Winner · 3/8 categoriesMistral Small 4 (Reasoning)· Qwen3.5-27B
Pick Qwen3.5-27B if you want the stronger benchmark profile. Mistral Small 4 (Reasoning) only becomes the better choice if its workflow or ecosystem matters more than the raw scoreboard.
Qwen3.5-27B is clearly ahead on the aggregate, 71 to 64. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Qwen3.5-27B's sharpest advantage is in multimodal & grounded, where it averages 75 against 60. The single biggest benchmark swing on the page is LiveCodeBench, 63.6% to 80.7%.
Qwen3.5-27B gives you the larger context window at 262K, compared with 256K for Mistral Small 4 (Reasoning).
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 | Mistral Small 4 (Reasoning) | Qwen3.5-27B |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 41.6% |
| BrowseComp | — | 61% |
| OSWorld-Verified | — | 56.2% |
| tau2-bench | — | 79% |
| CodingQwen3.5-27B wins | ||
| LiveCodeBench | 63.6% | 80.7% |
| SWE-bench Verified | — | 72.4% |
| Multimodal & GroundedQwen3.5-27B wins | ||
| MMMU-Pro | 60% | 75% |
| Reasoning | ||
| LongBench v2 | — | 60.6% |
| KnowledgeQwen3.5-27B wins | ||
| GPQA | 71.2% | 85.5% |
| MMLU-Pro | 78% | 86.1% |
| SuperGPQA | — | 65.6% |
| Instruction Following | ||
| IFEval | — | 95% |
| Multilingual | ||
| MMLU-ProX | — | 82.2% |
| Mathematics | ||
| AIME 2025 | 83.8% | — |
Qwen3.5-27B is ahead overall, 71 to 64. The biggest single separator in this matchup is LiveCodeBench, where the scores are 63.6% and 80.7%.
Qwen3.5-27B has the edge for knowledge tasks in this comparison, averaging 80.6 versus 75.6. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for coding in this comparison, averaging 77.6 versus 63.6. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for multimodal and grounded tasks in this comparison, averaging 75 versus 60. Inside this category, MMMU-Pro 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.