Qwen3.5 397B-A17B

Qwen3.5 397B-A17B ranks first on 5 of the 118 benchmarks it has scores on.

Qwen Released 13 Feb 2026 Family Qwen3.5 The lab's model card

124

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Qwen Feb 2026

Open weights Flagship

146.8 #77 of 271

Epoch Capabilities Index

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Benchmark	Score	Rank	Source
MMLU-Pro	87.8%	#6 of 151	Model card
MMLU-Redux	94.9%	#2 of 38	Model card
C-Eval	93%	#4 of 38	Model card
IFEval	92.6%	#14 of 126	Model card
BullshitBench	65.5%	#12 of 107	Official leaderboard
BullshitBench V2	78%	#14 of 122	Official leaderboard
Judgemark v2.1	85.5	#5 of 42	Official leaderboard
SEA-HELM	76.2	#8 of 62	Official leaderboard
HallusionBench	71.4%	#3 of 23	Model card
MMLU-ProX	84.7%	#3 of 17	Model card
NOVA-63	59.1%	#1 of 5	Model card
LMArena (Elo)	1,441	#74 of 333	Official leaderboard
MultiChallenge	67.6%	#8 of 36	Model card
LongBench v2	63.2%	#6 of 27	Model card
SuperGPQA	70.4%	#6 of 26	Model card
INCLUDE	85.6%	#6 of 25	Model card
IFBench	76.5%	#18 of 70	Model card
Humanity's Last Exam	48.3%	#42 of 147	Model card
HELM Arabic	80.2	#8 of 28	Official leaderboard
MMMLU	88.5%	#18 of 62	Model card
AA-LCR (Artificial Analysis Long Context Reasoning)	68.7%	#10 of 31	Model card
EQ-Bench 3	1,451	#28 of 78	Official leaderboard
Longform Creative Writing (EQ-Bench)	68.3	#46 of 119	Official leaderboard
HELM Arabic Enterprise	65.3	#12 of 23	Official leaderboard
SimpleVQA	67.1%	#12 of 23	Model card
WideSearch	74%	#9 of 17	Model card
Creative Writing v3	66.3	#64 of 120	Official leaderboard
Judgemark v4	54.2	#23 of 40	Official leaderboard
MAXIFE	88.2%	#3 of 5	Model card
Extended NYT Connections	58.9%	#61 of 97	Official leaderboard
Global PIQA	89.8%	#6 of 8	Model card
Seal-0	46.9%	#4 of 5	Model card
WMT24++	78.9%	#12 of 12	Model card

Benchmark	Score	Rank	Source
PolyMath	73.3%	#4 of 18	Model card
General365	57.7%	#6 of 23	Official leaderboard
HMMT 2025	94.8%	#13 of 49	Model card
GPQA Diamond	86.4%	#65 of 219	Epoch AI
OTIS Mock AIME 2024-2025	88.9%	#63 of 202	Epoch AI
DTBench	79.1%	#62 of 178	Official leaderboard
HMMT Nov 2025	92.7%	#11 of 23	Model card
HMMT 2026	87.9%	#17 of 31	Official leaderboard
CritPt	0.86%	#96 of 160	Epoch AI
IMO-AnswerBench	80.9%	#23 of 37	Model card
Echo General AI Prediction Leaderboard	986	#12 of 17	Official leaderboard
AIME 2026	94.2%	#22 of 31	Official leaderboard
FrontierMath Tiers 1-3 v2	31.2%	#69 of 89	Epoch AI
ResearchClawBench	14.2	#18 of 23	Official leaderboard
USAMO 2026	36.3%	#8 of 9	Official leaderboard

Benchmark	Score	Rank	Source
SWE-bench Verified	76.4%	#38 of 150	Model card
LiveCodeBench v6	83.6%	#23 of 76	Model card
SWE-bench Multilingual	69.3%	#43 of 70	Model card
Code Arena	1,400	#81 of 124	Official leaderboard
EvoCode-Bench	4.60%	#11 of 13	Official leaderboard

Benchmark	Score	Rank	Source
VITA-Bench	49.7%	#1 of 12	Model card
BFCL V4	72.9%	#3 of 24	Model card
tau3-bench	87.9%	#1 of 7	Official leaderboard
tau2-bench	86.7%	#23 of 69	Model card
BrowseComp-ZH	70.3%	#6 of 17	Model card
LisanBench	1293.3	#51 of 133	Official leaderboard
BrowseComp	78.6%	#30 of 77	Model card
Toolathlon	38.3%	#23 of 50	Official leaderboard
DeepPlanning	34.3%	#3 of 6	Model card
AndroidWorld	66.8%	#10 of 18	Model card
Terminal-Bench 2.0	52.5%	#34 of 61	Model card
SnakeBench	23	#105 of 183	Official leaderboard
Mystery Game Puzzles	18%	#51 of 80	Epoch AI
ScreenSpot-Pro	65.6%	#16 of 25	Model card
Claw-Eval Pass^3	56.8%	#15 of 23	Official leaderboard
Claw-Eval	74.2%	#16 of 23	Official leaderboard
MobileWorld	42.7%	#14 of 18	Official leaderboard
DeepTerminalBench	24%	#7 of 9	Official leaderboard
MCPMark	46.1%	#8 of 10	Model card
OSWorld-Verified	62.2%	#30 of 37	Model card
LiveClawBench	57%	#14 of 17	Official leaderboard
Pencil Puzzle Bench	0%	#47 of 53	Official leaderboard
Toolathlon-Verified	40.7%	#23 of 25	Official leaderboard
τ³-bench Banking	9.79%	#25 of 25	Official leaderboard
Claw-Eval-Live	49.5%	#12 of 12	Official leaderboard

Benchmark	Score	Rank	Source
MedXpertQA-MM	70%	#5 of 12	Model card
PMC-VQA	64.2%	#2 of 4	Model card
ProfBench	45.1%	#44 of 72	Official leaderboard
SpreadsheetBench 2	11.2%	#7 of 8	Official leaderboard
APEX-Agents	24.9%	#53 of 56	Epoch AI
SLAKE	79.9%	#4 of 4	Model card

Benchmark	Score	Rank	Source
MathVista	90.3%	#2 of 64	Model card
MMMU	85%	#4 of 97	Model card
MMBench-EN v1.1	93.7%	#1 of 17	Model card
MMStar	83.8%	#2 of 30	Model card
OCRBench	93.1%	#2 of 28	Model card
DynaMath	86.3%	#2 of 17	Model card
RefSpatial-Bench	73.6%	#1 of 8	Model card
CountBench	97.2%	#3 of 20	Model card
VideoMME v2 (with subtitles)	39.1%	#5 of 31	Official leaderboard
We-Math	87.9%	#2 of 11	Model card
MVBench	77.6%	#3 of 15	Model card
AI2D	93.9%	#9 of 44	Model card
MMMU-Pro	79%	#15 of 72	Model card
MMLongBench-Doc	61.5%	#3 of 13	Model card
MATH-Vision (MathVision)	88.6%	#10 of 42	Model card
V*	95.8%	#3 of 12	Model card
Video-MME	87.5%	#11 of 42	Model card
Video-MMMU	84.7%	#9 of 30	Model card
EmbSpatialBench	84.5%	#3 of 10	Model card
MLVU	86.7%	#5 of 16	Model card
ERQA	67.5%	#9 of 26	Model card
MMVU	75.4%	#7 of 20	Model card
Vision Arena	1,246	#49 of 136	Official leaderboard
CC-OCR	82%	#3 of 8	Model card
SUNRGBD	38.3	#2 of 5	Model card
RealWorldQA	83.9%	#13 of 30	Model card
CharXiv	80.8%	#30 of 67	Model card
Nuscene	16.0	#1 of 2	Model card
LVBench	75.5%	#13 of 25	Model card
BabyVision	43.3%	#15 of 27	Official leaderboard
ODinW-13	47%	#5 of 9	Model card
Claw-Eval-MM	50.2%	#7 of 11	Official leaderboard
Hypersim	12.5	#2 of 3	Model card
LingoQA	81.6%	#3 of 4	Model card

Benchmark	Score		Rank	Source
SMDD-Bench	27.5%		#5 of 7	Official leaderboard
SecCodeBench	68.3%		#1 of 1	Model card

Benchmark	Score	Rank	Source
LMCA	37.9%	#58 of 139	Epoch AI
Conceptual Reasoning Index	49.7	#68 of 149	Official leaderboard
SAGE v1.1 (Vals)	85.5%	#14 of 28	Official leaderboard
ACCoRD	55.6	#102 of 150	Official leaderboard

Qwen's models in color, against every other model on the Epoch Capabilities Index, by release date.

The benchmarks the lab chose to headline, with the scores it gave.