Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
DeepSeek V3.1 (Reasoning)
43
0/8 categoriesQwen3.5-27B
71
Winner · 7/8 categoriesDeepSeek V3.1 (Reasoning)· Qwen3.5-27B
Pick Qwen3.5-27B if you want the stronger benchmark profile. DeepSeek V3.1 (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 43. 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 coding, where it averages 77.6 against 19. The single biggest benchmark swing on the page is LiveCodeBench, 16% to 80.7%.
Qwen3.5-27B gives you the larger context window at 262K, compared with 128K for DeepSeek V3.1 (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 | DeepSeek V3.1 (Reasoning) | Qwen3.5-27B |
|---|---|---|
| AgenticQwen3.5-27B wins | ||
| Terminal-Bench 2.0 | 42% | 41.6% |
| BrowseComp | 48% | 61% |
| OSWorld-Verified | 44% | 56.2% |
| tau2-bench | — | 79% |
| CodingQwen3.5-27B wins | ||
| HumanEval | 26% | — |
| SWE-bench Verified | 14% | 72.4% |
| LiveCodeBench | 16% | 80.7% |
| SWE-bench Pro | 25% | — |
| Multimodal & GroundedQwen3.5-27B wins | ||
| MMMU-Pro | 37% | 75% |
| OfficeQA Pro | 47% | — |
| ReasoningQwen3.5-27B wins | ||
| MuSR | 30% | — |
| BBH | 64% | — |
| LongBench v2 | 57% | 60.6% |
| MRCRv2 | 56% | — |
| KnowledgeQwen3.5-27B wins | ||
| MMLU | 34% | — |
| GPQA | 33% | 85.5% |
| SuperGPQA | 31% | 65.6% |
| MMLU-Pro | 53% | 86.1% |
| HLE | 10% | — |
| FrontierScience | 37% | — |
| SimpleQA | 32% | — |
| Instruction FollowingQwen3.5-27B wins | ||
| IFEval | 70% | 95% |
| MultilingualQwen3.5-27B wins | ||
| MGSM | 64% | — |
| MMLU-ProX | 61% | 82.2% |
| Mathematics | ||
| AIME 2023 | 34% | — |
| AIME 2024 | 36% | — |
| AIME 2025 | 35% | — |
| HMMT Feb 2023 | 30% | — |
| HMMT Feb 2024 | 32% | — |
| HMMT Feb 2025 | 31% | — |
| BRUMO 2025 | 33% | — |
| MATH-500 | 62% | — |
Qwen3.5-27B is ahead overall, 71 to 43. The biggest single separator in this matchup is LiveCodeBench, where the scores are 16% and 80.7%.
Qwen3.5-27B has the edge for knowledge tasks in this comparison, averaging 80.6 versus 32.5. 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 19. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for reasoning in this comparison, averaging 60.6 versus 49.5. Inside this category, LongBench v2 is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for agentic tasks in this comparison, averaging 51.6 versus 44.2. Inside this category, BrowseComp 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 41.5. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for instruction following in this comparison, averaging 95 versus 70. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Qwen3.5-27B has the edge for multilingual tasks in this comparison, averaging 82.2 versus 62.1. Inside this category, MMLU-ProX 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.