Noometry

Alibaba (Qwen), open weights

Qwen3.5 397B-A17B

Qwen3.5 397B-A17B by Alibaba (Qwen) ranks 67th of 354 ranked models on the Noometry Index as of October 2026, with a score of 46.0. Its strongest category is multimodal, where it ranks 44th. API pricing starts at $0.60 per million input tokens and $3.60 per million output tokens, with a 262K-token context window.

Last verified

Noometry rank
#67 of 354
Index score
46.0
Evidence
Confirmed 36 results
Provider
Alibaba (Qwen)
Released
February 1, 2026
Weights
Open weights
Reasoning
Yes
Context window
262K
Max output
66K
Input price
$0.60 / M
Output price
$3.60 / M
Blended price
$1.35 / M
Output speed
9 tokens/s Kagi
Value
#128 of 219
Knowledge cutoff
January 2025
Input
text, image, video, audio
Hugging Face
Qwen/Qwen3.5-397B-A17B

Category scores

Each category score combines every public result we have in that category.

  1. Coding 42.0
  2. Agentic & Tool Use 33.3
  3. Reasoning 34.5
  4. Math 46.1
  5. Knowledge 53.3
  6. Multimodal 40.7
  7. Multilingual 53.7
  8. Instruction Following 75.0
  9. Long Context 44.1
  10. Writing & Preference 62.3
CategoryScoreRankResults
Coding42.0#1142
Agentic & Tool Use33.3#535
Reasoning34.5#708
Math46.1#733
Knowledge53.3#582
Multimodal40.7#441
Multilingual53.7#591
Instruction Following75.0#771
Long Context44.1#741
Writing & Preference62.3#794

Strengths and weaknesses

Categories where Qwen3.5 397B-A17B places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Knowledge53.3+16.0#58 of 314, top 19%
Multilingual53.7+6.3#59 of 297, top 20%
Reasoning34.5+10.9#70 of 350, top 20%

Weakest categories

CategoryScorevs medianRank
Agentic & Tool Use33.3+3.0#53 of 154, top 35%
Multimodal40.7+2.2#44 of 128, top 35%
Coding42.0+3.3#114 of 340, top 34%

Closest competitors

The models ranked just above and below Qwen3.5 397B-A17B. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
Inkling-Small#6346.5$0.64—Compare
GPT-5 Pro#6446.4$41.255Compare
Grok 4.20 Multi-Agent#6546.2$1.56—Compare
GLM-5#6646.1$1.5523Compare
Qwen3.8 27B#6846.0$1.11—Compare
GPT-5.3 Codex#6945.8$4.81—Compare
Kimi K2 Thinking Turbo#7045.8——Compare
Qwen3.5 Max Preview#7145.3——Compare

Sponsored placements are available on pages like this one. Advertise on Noometry

Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

Coding

BenchmarkScorePositionSettingSourceDate
LMArena WebDev1400#72 of 113, top 64%LMArena2026-10-08
LMArena Coding1465#65 of 294, top 23%LMArena2026-10-08

Agentic & Tool Use

BenchmarkScorePositionSettingSourceDate
APEX-Agents24.9%#46 of 49, top 94%Epoch AI
τ²-bench Airline81.5%#5 of 7, top 72%enabledτ²-bench2026-03-02
τ²-bench Banking9.8%#26 of 26, top 100%enabledτ²-bench2026-03-02
τ²-bench Retail84.4%Best of 7enabledτ²-bench2026-03-02
τ²-bench Telecom97.8%Best of 7enabledτ²-bench2026-03-02

Reasoning

BenchmarkScorePositionSettingSourceDate
Kagi LLM Benchmark73.7%#15 of 99, top 16%Kagi LLM Benchmark
NYT Connections (extended)58.9%#58 of 91, top 64%Lech Mazur benchmarks
Chess Puzzles12%Epoch AI2026-08-07
Chess Puzzles13%#74 of 129, top 58%noneEpoch AI2026-08-07
Thematic Generalization65.1%#8 of 23, top 35%Lech Mazur benchmarks
LMArena Hard Prompts1448#65 of 297, top 22%LMArena2026-10-08
Mystery Game Puzzles18%#48 of 74, top 65%noneEpoch AI2026-08-27
DTBench87.5%#51 of 151, top 34%Epoch AI
LMCA37.9%#52 of 125, top 42%Epoch AI
Epoch Capabilities Index146.65#71 of 213, top 34%Epoch AI2026-02-13

Math

BenchmarkScorePositionSettingSourceDate
FrontierMath (Tiers 1-3)29.5%Epoch AI2026-08-28
FrontierMath (Tiers 1-3)31.2%#61 of 81, top 76%noneEpoch AI2026-08-30
OTIS Mock AIME 2024-202588.9%#57 of 173, top 33%Epoch AI2026-08-07
OTIS Mock AIME 2024-202582.2%noneEpoch AI2026-08-07
LMArena Math1454#60 of 285, top 22%LMArena2026-10-08

Knowledge

BenchmarkScorePositionSettingSourceDate
GPQA Diamond85.9%Epoch AI2026-08-07
GPQA Diamond86.4%#58 of 186, top 32%noneEpoch AI2026-08-07
LMArena Expert1462#60 of 273, top 22%LMArena2026-10-08

Multimodal

BenchmarkScorePositionSettingSourceDate
LMArena Vision1263#46 of 122, top 38%LMArena2026-10-09

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1430#59 of 297, top 20%LMArena2026-10-08
LMArena Chinese1500#44 of 285, top 16%LMArena2026-10-08
LMArena French1461#46 of 223, top 21%LMArena2026-10-08
LMArena German1447#47 of 231, top 21%LMArena2026-10-08
LMArena Japanese1426#36 of 211, top 18%LMArena2026-10-08
LMArena Korean1384#73 of 213, top 35%LMArena2026-10-08
LMArena Russian1429#69 of 283, top 25%LMArena2026-10-08
LMArena Spanish1441#65 of 226, top 29%LMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1424#69 of 298, top 24%LMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
LMArena Longer Query1442#61 of 291, top 21%LMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1438#67 of 297, top 23%LMArena2026-10-08
LMArena Creative Writing1401#80 of 295, top 28%LMArena2026-10-08
EQ-Bench Creative Writing1478#60 of 115, top 53%EQ-Bench
LMArena Multi-Turn1446#60 of 295, top 21%LMArena2026-10-08

API pricing by provider

RouteInput $/MOutput $/MCached input $/MChecked
alibaba$0.60$3.60—2026-10-10
deepinfra$0.45$3$0.222026-10-10
openrouter$0.55$3.50$0.232026-10-10
together$0.60$3.60$0.352026-10-10

All Alibaba (Qwen) API prices → Estimate your cost →

Compare Qwen3.5 397B-A17B

Other Alibaba (Qwen) models

Frequently asked questions

How good is Qwen3.5 397B-A17B?

Qwen3.5 397B-A17B by Alibaba (Qwen) ranks 67th of 354 ranked models on the Noometry Index as of October 2026, with a score of 46.0. Its strongest category is multimodal, where it ranks 44th. API pricing starts at $0.60 per million input tokens and $3.60 per million output tokens, with a 262K-token context window.

How much does Qwen3.5 397B-A17B cost?

Qwen3.5 397B-A17B costs $0.60 per million input tokens and $3.60 per million output tokens on Alibaba (Qwen)'s own API.

What is Qwen3.5 397B-A17B's context window?

Qwen3.5 397B-A17B accepts up to 262K tokens of input and can write up to 66K tokens in one response.

Is Qwen3.5 397B-A17B open source?

Yes. Qwen3.5 397B-A17B's weights are downloadable from Hugging Face (Qwen/Qwen3.5-397B-A17B); check the license for commercial terms.

How fast is Qwen3.5 397B-A17B?

Qwen3.5 397B-A17B generated about 9 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.

What are Qwen3.5 397B-A17B's strengths and weaknesses?

Relative to other ranked models, Qwen3.5 397B-A17B places best in knowledge, multilingual, reasoning and lowest in agentic & tool use, multimodal, coding.

What is Qwen3.5 397B-A17B best at?

Its best category is multimodal, where it ranks 44th on Noometry.