Model card
Qwen3.5-397B-A17B.
Alibabaopen-source
§ 02 · Benchmarks
Every benchmark Qwen3.5-397B-A17B has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | MMMU | Multimodal · Image-Text-to-Text | accuracy | 85.0% | #1 | — | source ↗ |
| 02 | MMStar | Multimodal · Image-Text-to-Text | accuracy | 83.8% | #1 | — | source ↗ |
| 03 | MVBench | Multimodal · Video Understanding | accuracy | 77.6% | #2 | — | source ↗ |
| 04 | Video-MME | Multimodal · Video Understanding | accuracy | 87.5% | #2 | — | source ↗ |
| 05 | Tau2-Bench | Agentic AI · Tool Use | accuracy | 86.7% | #3 | — | source ↗ |
| 06 | BrowseComp | Natural Language Processing · Question Answering | accuracy | 69.0% | #5 | — | source ↗ |
| 07 | MMMU-Pro | Multimodal · Visual Question Answering | accuracy | 79.0% | #5 | — | source ↗ |
| 08 | OmniDocBench | Computer Vision · Document Parsing | accuracy | 90.8% | #6 | — | source ↗ |
| 09 | RealWorldQA | Multimodal · Visual Question Answering | accuracy | 83.9% | #6 | — | source ↗ |
| 10 | LiveCodeBench | Computer Code · Code Generation | pass-1 | 83.6% | #7 | — | source ↗ |
| 11 | GPQA Diamond | Reasoning · Multi-step Reasoning | accuracy | 88.4% | #8 | — | source ↗ |
| 12 | SWE-Bench Verified | Computer Code · Code Generation | accuracy | 76.4% | #9 | — | source ↗ |
| 13 | HLE | Reasoning · Multi-step Reasoning | accuracy | 28.7% | #18 | — | source ↗ |
| 14 | PLCC | Natural Language Processing · Polish Cultural Competency | grammar | 76.0% | #30 | — | source ↗ |
| 15 | PLCC | Natural Language Processing · Polish Cultural Competency | geography | 85.0% | #41 | — | source ↗ |
| 16 | PLCC | Natural Language Processing · Polish Cultural Competency | vocabulary | 70.0% | #43 | — | source ↗ |
| 17 | PLCC | Natural Language Processing · Polish Cultural Competency | average | 75.0% | #46 | — | source ↗ |
| 18 | PLCC | Natural Language Processing · Polish Cultural Competency | history | 83.0% | #49 | — | source ↗ |
| 19 | PLCC | Natural Language Processing · Polish Cultural Competency | art-and-entertainment | 63.0% | #54 | — | source ↗ |
| 20 | PLCC | Natural Language Processing · Polish Cultural Competency | culture-and-tradition | 73.0% | #60 | — | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area
Where Qwen3.5-397B-A17B actually performs.
§ 04 · Papers
1 paper with results for Qwen3.5-397B-A17B.
- 2026-02-16· 13 results
Qwen3.5: Towards Native Multimodal Agents
§ 05 · Related models
Other Alibaba models scored on Codesota.
§ 06 · Sources & freshness
Where these numbers come from.
pwc-dump
13
results
sdadas/PLCC
7
results
7 of 20 rows marked verified.