Qwen 3.5
Alibaba Cloud's Qwen 3.5 is a practical candidate for coding agents and structured automation in MyClaw. It helps you choose between the 9B, 27B, and 35B options, plan benchmarks, and decide when to test a move to Qwen 3.6.
QwenWhere Qwen 3.5 fits
Use this page as a practical model-selection guide when evaluating Qwen 3.5 for coding agents, structured workflows, and migration planning.
Agentic task execution
Qwen is positioned around tool use and MCP-style workflows, which makes Qwen 3.5 a useful candidate for agents that need to plan, call tools, inspect files, and keep context across steps.
Model-size selection
Start with Qwen 3.5 9B for lightweight routing or local-style tasks, move to Qwen 3.5 27B for balanced work, and reserve Qwen 3.5 35B for harder coding and agent sessions.
Benchmark-first rollout
Treat Qwen 3.5 benchmarks as workload signals: compare coding, math, instruction following, multilingual behavior, tool reliability, and cost before promoting a workflow to production.
Qwen 3.5 Benchmark results
Official Qwen results compare Qwen3.5-397B-A17B with leading frontier models across instruction following, reasoning, tool use, search, coding, multilingual knowledge, document understanding, and video reasoning benchmarks.
Language benchmark results
Official Qwen3.5 results compare Qwen3.5-397B-A17B with GPT-5.2, Claude 4.5 Opus, Gemini-3 Pro, Qwen3-Max-Thinking, and K2.5-1T-A32B across knowledge, instruction following, long-context, and multilingual benchmarks.
| GPT5.2 | Claude 4.5 Opus | Gemini-3 Pro | Qwen3-Max-Thinking | K2.5-1T-A32B | Qwen3.5-397B-A17B | |
|---|---|---|---|---|---|---|
| Knowledge | ||||||
| MMLU-Pro | 87.4 | 89.5 | 89.8 | 85.7 | 87.1 | 87.8 |
| MMLU-Redux | 95.0 | 95.6 | 95.9 | 92.8 | 94.5 | 94.9 |
| SuperGPQA | 67.9 | 70.6 | 74.0 | 67.3 | 69.2 | 70.4 |
| C-Eval | 90.5 | 92.2 | 93.4 | 93.7 | 94.0 | 93.0 |
| Instruction Following | ||||||
| IFEval | 94.8 | 90.9 | 93.5 | 93.4 | 93.9 | 92.6 |
| IFBench | 75.4 | 58.0 | 70.4 | 70.9 | 70.2 | 76.5 |
| MultiChallenge | 57.9 | 54.2 | 64.2 | 63.3 | 62.7 | 67.6 |
| Long Context | ||||||
| AA-LCR | 72.7 | 74.0 | 70.7 | 68.7 | 70.0 | 68.7 |
| LongBench v2 | 54.5 | 64.4 | 68.2 | 60.6 | 61.0 | 63.2 |
| Multilingualism | ||||||
| MMMLU | 89.5 | 90.1 | 90.6 | 84.4 | 86.0 | 88.5 |
| MMLU-ProX | 83.7 | 85.7 | 87.7 | 78.5 | 82.3 | 84.7 |
| NOVA-63 | 54.6 | 56.7 | 56.7 | 54.2 | 56.0 | 59.1 |
| INCLUDE | 87.5 | 86.2 | 90.5 | 82.3 | 83.3 | 85.6 |
| Global PIQA | 90.9 | 91.6 | 93.2 | 86.0 | 89.3 | 89.8 |
| PolyMATH | 62.5 | 79.0 | 81.6 | 64.7 | 43.1 | 73.3 |
| WMT24++ | 78.8 | 79.7 | 80.7 | 77.6 | 77.6 | 78.9 |
| MAXIFE | 88.4 | 79.2 | 87.5 | 84.0 | 72.8 | 88.2 |
Vision Language benchmark results
Official Qwen3.5 vision-language results compare Qwen3.5-397B-A17B with GPT-5.2, Claude 4.5 Opus, Gemini-3 Pro, Qwen3-VL-235B-A22B, and K2.5-1T-A32B across STEM, VQA, document understanding, spatial reasoning, video, visual-agent, and medical VQA tasks.
| GPT5.2 | Claude 4.5 Opus | Gemini-3 Pro | Qwen3-VL-235B-A22B | K2.5-1T-A32B | Qwen3.5-397B-A17B | |
|---|---|---|---|---|---|---|
| STEM and Puzzle | ||||||
| MMMU | 86.7 | 80.7 | 87.2 | 80.6 | 84.3 | 85.0 |
| MMMU-Pro | 79.5 | 70.6 | 81.0 | 69.3 | 78.5 | 79.0 |
| MathVision | 83.0 | 74.3 | 86.6 | 74.6 | 84.2 | 88.6 |
| Mathvista(mini) | 83.1 | 80.0 | 87.9 | 85.8 | 90.1 | 90.3 |
| We-Math | 79.0 | 70.0 | 86.9 | 74.8 | 84.7 | 87.9 |
| DynaMath | 86.8 | 79.7 | 85.1 | 82.8 | 84.4 | 86.3 |
| ZEROBench | 9 | 3 | 10 | 4 | 9 | 12 |
| ZEROBench_sub | 33.2 | 28.4 | 39.0 | 28.4 | 33.5 | 41.0 |
| BabyVision | 34.4 | 14.2 | 49.7 | 22.2 | 36.5 | 52.3/43.3 |
| General VQA | ||||||
| RealWorldQA | 83.3 | 77.0 | 83.3 | 81.3 | 81.0 | 83.9 |
| MMStar | 77.1 | 73.2 | 83.1 | 78.7 | 80.5 | 83.8 |
| HallusionBench | 65.2 | 64.1 | 68.6 | 66.7 | 69.8 | 71.4 |
| MMBenchEN-DEV-v1.1 | 88.2 | 89.2 | 93.7 | 89.7 | 94.2 | 93.7 |
| SimpleVQA | 55.8 | 65.7 | 73.2 | 61.3 | 71.2 | 67.1 |
| Text Recognition and Document Understanding | ||||||
| OmniDocBench1.5 | 85.7 | 87.7 | 88.5 | 84.5 | 88.8 | 90.8 |
| CharXiv(RQ) | 82.1 | 68.5 | 81.4 | 66.1 | 77.5 | 80.8 |
| MMLongBench-Doc | -- | 61.9 | 60.5 | 56.2 | 58.5 | 61.5 |
| CC-OCR | 70.3 | 76.9 | 79.0 | 81.5 | 79.7 | 82.0 |
| AI2D_TEST | 92.2 | 87.7 | 94.1 | 89.2 | 90.8 | 93.9 |
| OCRBench | 80.7 | 85.8 | 90.4 | 87.5 | 92.3 | 93.1 |
| Spatial Intelligence | ||||||
| ERQA | 59.8 | 46.8 | 70.5 | 52.5 | -- | 67.5 |
| CountBench | 91.9 | 90.6 | 97.3 | 93.7 | 94.1 | 97.2 |
| RefCOCO(avg) | -- | -- | 84.1 | 91.1 | 87.8 | 92.3 |
| ODInW13 | -- | -- | 46.3 | 43.2 | -- | 47.0 |
| EmbSpatialBench | 81.3 | 75.7 | 61.2 | 84.3 | 77.4 | 84.5 |
| RefSpatialBench | -- | -- | 65.5 | 69.9 | -- | 73.6 |
| LingoQA | 68.8 | 78.8 | 72.8 | 66.8 | 68.2 | 81.6 |
| V* | 75.9 | 67.0 | 88.0 | 85.9 | 77.0 | 95.8/91.1 |
| Hypersim | -- | -- | -- | 11.0 | -- | 12.5 |
| SUNRGBD | -- | -- | -- | 34.9 | -- | 38.3 |
| Nuscene | -- | -- | -- | 13.9 | -- | 16.0 |
| Video Understanding | ||||||
| VideoMME(w sub.) | 86 | 77.6 | 88.4 | 83.8 | 87.4 | 87.5 |
| VideoMME(w/o sub.) | 85.8 | 81.4 | 87.7 | 79.0 | 83.2 | 83.7 |
| VideoMMMU | 85.9 | 84.4 | 87.6 | 80.0 | 86.6 | 84.7 |
| MLVU (M-Avg) | 85.6 | 81.7 | 83.0 | 83.8 | 85.0 | 86.7 |
| MVBench | 78.1 | 67.2 | 74.1 | 75.2 | 73.5 | 77.6 |
| LVBench | 73.7 | 57.3 | 76.2 | 63.6 | 75.9 | 75.5 |
| MMVU | 80.8 | 77.3 | 77.5 | 71.1 | 80.4 | 75.4 |
| Visual Agent | ||||||
| ScreenSpot Pro | -- | 45.7 | 72.7 | 62.0 | -- | 65.6 |
| OSWorld-Verified | 38.2 | 66.3 | -- | 38.1 | 63.3 | 62.2 |
| AndroidWorld | -- | -- | -- | 63.7 | -- | 66.8 |
| Medical VQA | ||||||
| SLAKE | 76.9 | 76.4 | 81.3 | 72.5 | 81.6 | 79.9 |
| PMC-VQA | 58.9 | 59.9 | 62.3 | 56.1 | 63.3 | 64.2 |
| MedXpertQA-MM | 73.3 | 63.6 | 76.0 | 47.6 | 65.3 | 70.0 |
Qwen 3.5 vs 3.6
For most teams, Qwen 3.5 vs 3.6 is not a branding choice. It is a migration test across benchmark scores, latency, tool reliability, model availability, and deployment cost.
| Model comparison | Qwen 3.5 | Qwen 3.6 |
|---|---|---|
| First evaluation | Use Qwen 3.5 as the stable baseline for benchmark runs, model-size selection, and agent workflows already tuned around 9B, 27B, or 35B options. | Use Qwen 3.6 as the challenger when you want to test newer capabilities or a newer open/proprietary Qwen release path. |
| Benchmark focus | Prioritize coding, tool calls, JSON output discipline, multilingual tasks, and latency by model size. | Run the same suite before migration, then compare win rate, regression count, output length, and retry behavior. |
| Model-size planning | The 9B, 27B, and 35B framing makes Qwen 3.5 useful for staged rollout from lightweight tasks to harder agent sessions. | Check which Qwen 3.6 sizes are actually available in your target provider before assuming a one-to-one replacement. |
| Best first test | Start with Qwen 3.5 when you need a controlled benchmark page and predictable model-selection ladder. | Test Qwen 3.6 when the workload needs newer Qwen behavior and you can tolerate migration validation. |
Run Qwen 3.5 vs 3.6 on the same prompts, tools, and scoring harness; do not compare isolated leaderboard numbers without matching settings.
Where Qwen 3.5 fits in MyClaw
Repository agents
Use Qwen 3.5 to inspect code, propose changes, and run implementation loops inside a hosted MyClaw workspace.
Structured data tasks
Evaluate JSON output, table reasoning, and deterministic extraction before using the model in operational workflows.
Multilingual support
Test customer-facing or internal workflows across the languages your agents actually need to serve.
Migration checks
Compare Qwen 3.5 vs 3.6 with the same tool schemas, context size, retry policy, and acceptance tests.
Qwen 3.5 FAQs
Qwen 3.5
Run Qwen 3.5 in a hosted MyClaw agent
Use MyClaw to benchmark Qwen 3.5 with real files, tools, coding tasks, multilingual prompts, and Qwen 3.5 vs 3.6 migration checks.