Qwen3.5-397B-A17B

AlibabaOpenApache 2.0

Qwen3.5-397B-A17B is a vision-language model from Alibaba. It ranks 54 of 167 on LMArena’s overall vision board with 1262±5 from 32,657 votes. Its best category is Entity recognition (#14 of 51). License: Apache 2.0.

Key figures

LMArena · Overall1262±5#54 / 167 · Votes 32,657
OCR1271±5#51 / 123 · Votes 22,696
Diagrams1277±7#52 / 123 · Votes 8,623
Homework1292±10#45 / 118 · Votes 4,001
Captioning1230±57#15 / 34 · Votes 78
Entity recognition1265±37#14 / 51 · Votes 222
Creative writing1260±13#44 / 104 · Votes 2,304
Humor1262±17#42 / 93 · Votes 1,223
English1268±7#51 / 167 · Votes 12,659
Chinese1314±15#43 / 128 · Votes 2,045
Developer
Alibaba
License
Apache 2.0
First seen (our snapshots)
2026-10-26

Who it suits

  • Overall #54 of 167: upper half.
  • Strongest category: Entity recognition (#14 of 51).
  • Open weights (Apache 2.0): you can self-host it.

Neighbours on the overall board

#ModelScoreVotesLicenseΔ vs this model
51Gemini 2.5 Pro Google126388,755Proprietary+1
52Grok 4.6 (High) xAI12637,261Proprietary+1
53MiMo-V2.6-Pro Xiaomi12631,754MIT+1
54Qwen3.5-397B-A17B Alibaba126232,657Apache 2.0
55Grok 4.20 Beta (Reasoning, 0309) xAI126226,796Proprietary+0
56Gemma 4 26B A4B Google126023,922Apache 2.0-2
57Grok 4.20 Multi-Agent Beta (0309) xAI125924,532Proprietary-3

Alternatives · More from Alibaba

#ModelScoreVotesLicense
10Qwen3.8-Max Alibaba13139,957Proprietary
37Qwen3.7-Plus Alibaba127814,212Proprietary
44Qwen3.8-27B Alibaba12697,614Apache 2.0
67Qwen3-VL-235B-A22B-Instruct Alibaba124712,724Apache 2.0
70Qwen3.5-122B-A10B Alibaba124413,236Apache 2.0

Alternatives · Top open-weight models

#ModelScoreVotesLicense
24GLM-5.3-Flash Z.ai129416,254MIT
34Kimi K2.6 Moonshot AI128216,368Modified MIT
38DeepSeek-V4.1-Flash-20260910 (Max) DeepSeek12784,100MIT
39Gemma 4 31B Google127640,219Apache 2.0
44Qwen3.8-27B Alibaba12697,614Apache 2.0

How to read these scores

  • If two scores differ by less than their ± intervals, the models are statistically close.
  • Each category board is computed independently; do not compare scores across categories.
  • Models with few votes move more; ranks of new entries may change.

FAQ

Where does Qwen3.5-397B-A17B rank among vision models?

Overall rank 54 (1262); OCR rank 51 (1271); Diagrams rank 52 (1277); Homework rank 45 (1292); Captioning rank 15 (1230); Entity recognition rank 14 (1265); Creative writing rank 44 (1260); Humor rank 42 (1262); English rank 51 (1268); Chinese rank 43 (1314) (LMArena, data as of Oct 26, 2026).

How good is Qwen3.5-397B-A17B at OCR?

Qwen3.5-397B-A17B ranks 51 of 123 on LMArena’s OCR board with 1271 (22,696 votes).

Is Qwen3.5-397B-A17B better with Chinese or English prompts?

Chinese board: #43 of 128 (1314); English board: #51 of 167 (1268). The boards are computed separately, so compare ranks rather than scores.

Is Qwen3.5-397B-A17B open-weight?

LMArena lists its license as Apache 2.0, not proprietary; check the license text for commercial use.

How far is Qwen3.5-397B-A17B from #1?

#1 overall is Claude Fable 5 (High) (1325); Qwen3.5-397B-A17B has 1262, 63 points behind.

Sources

LMArena data published Oct 19, 2026.

Synced here Oct 26, 2026.

Price, context length and release date have no reliable public data here and are not shown; all figures come from snapshots, blanks are –.