Qwen 3.8 27B

Qwen 3.8 27B places in the top three on AgentIF (2026-10 setting).

Qwen Released 14 Aug 2026

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Qwen Aug 2026

Open weights

151.8 #55 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
BullshitBench	76.4%	#8 of 107	Official leaderboard
AgentIF (2026-10 setting)	65.8%	#2 of 24	Official leaderboard
BullshitBench V2	79%	#13 of 122	Official leaderboard
IFBench	79.5%	#11 of 70	Model card
LMArena (Elo)	1,436	#83 of 333	Official leaderboard
Judgemark v4	67.4	#17 of 40	Official leaderboard
Humanity's Last Exam	30.8%	#73 of 147	Model card
Creative Writing v3	64.8	#71 of 120	Official leaderboard
Extended NYT Connections	54.5%	#63 of 97	Official leaderboard
Longform Creative Writing (EQ-Bench)	53.3	#79 of 119	Official leaderboard

Benchmark	Score	Rank	Source
GPQA Diamond	89.2%	#38 of 284	Model card
DTBench	80%	#57 of 178	Official leaderboard
SimpleBench	60.2%	#38 of 100	Official leaderboard
ARC-AGI	87.5%	#39 of 92	Official leaderboard
CritPt	5.43%	#69 of 160	Epoch AI
ARC-AGI-2	42.4%	#43 of 95	Official leaderboard
Surface Evolver Bench	45%	#18 of 26	Epoch AI
ProofBench	16%	#61 of 83	Epoch AI
ProofBench v1.1	16%	#45 of 51	Official leaderboard

Benchmark	Score	Rank	Source
LiveCodeBench v6	90.3%	#6 of 76	Model card
WebDev Arena	1,593	#20 of 105	Epoch AI
Code Arena	1,590	#29 of 124	Official leaderboard
SWE-Bench Pro	61.7%	#23 of 77	Model card
Image-to-WebDev Arena	1,577	#23 of 55	Official leaderboard
Vibe Code Bench	64.8%	#50 of 115	Official leaderboard
SWE-bench Multimodal	38.6%	#6 of 10	Model card
NL2Repo-Bench	42.3%	#19 of 28	Model card
Code Migration	14.2%	#58 of 78	Official leaderboard
Vision2Web	62.9%	#4 of 5	Model card
IOI v2 (Vals)	39.1%	#37 of 45	Official leaderboard
DeepSWE	42.2%	#47 of 51	Model card
QwenSWEBench	79%	#2 of 2	Model card

Benchmark	Score	Rank	Source
OSWorld-Verified	84.3%	#4 of 37	Model card
LLM Chess	846	#27 of 124	Official leaderboard
AndroidWorld	81.9%	#4 of 18	Model card
CivBench	1,516	#7 of 18	Official leaderboard
Terminal-Bench 2.1	73%	#32 of 53	Model card
WebArena-Verified	64.8%	#4 of 6	Model card
Agent Arena	-0.04	#38 of 53	Official leaderboard
WitnessBench	3.10%	#10 of 12	Official leaderboard
RecreationBench	47.1%	#3 of 3	Model card

Benchmark	Score	Rank	Source
Harvey LAB (Legal Agent Benchmark)	11.3%	#15 of 80	Official leaderboard
Tax Agent Bench	30.9%	#27 of 79	Official leaderboard
MedScribe	83.8%	#39 of 98	Official leaderboard
Excel Modeling Benchmark	59.7%	#36 of 76	Official leaderboard
MortgageTax	64.9%	#42 of 86	Official leaderboard
Legal Research Bench	36.1%	#44 of 88	Official leaderboard
LegalBench	82.4%	#70 of 138	Official leaderboard
APEX-Agents	47.5%	#34 of 56	Epoch AI
Vals Finance Agent v2	48.5%	#49 of 80	Official leaderboard
TaxEval v2	70.8%	#80 of 129	Official leaderboard
CoWorkBench	70.7%	#4 of 6	Model card
BlenderWorld	47.7	#10 of 13	Official leaderboard
MedCode	28.7%	#91 of 96	Official leaderboard
Agents' Last Exam	20.4%	#24 of 25	Model card
JobBench	33.4%	#12 of 12	Model card

Benchmark	Score	Rank	Source
OCRBench v2	63.5%	#6 of 65	Official leaderboard
MATH-Vision (MathVision)	94.6%	#4 of 42	Model card
CharXiv	90.2%	#7 of 67	Model card
RealWorldQA	85.9%	#7 of 30	Model card
ParseBench	69.7%	#12 of 42	Official leaderboard
BabyVision	85.6%	#7 of 20	Model card
Vision Arena	1,242	#53 of 136	Official leaderboard
ERQA	65.5%	#12 of 26	Model card
Claw-Eval-MM	56.9%	#4 of 7	Model card
GeoBench	38%	#45 of 48	Official leaderboard

Benchmark	Score		Rank	Source
CyberBench v1.1 (Vals)	67.8%		#24 of 43	Official leaderboard
Human Pathogen Capabilities Test v2 (HPCT v2)	48.6%		#47 of 71	Official leaderboard

Benchmark	Score	Rank	Source
SAGE (Vals)	52.4%	#7 of 81	Official leaderboard
ACCoRD	73.3	#32 of 150	Official leaderboard
Conceptual Reasoning Index	55.5	#51 of 149	Official leaderboard
LMCA	41.4%	#48 of 139	Epoch AI
Decision Index	57.0	#2 of 3	Official leaderboard
Political Science LLM Benchmark	63.5%	#23 of 29	Official leaderboard
LiveBench	58.4%	#24 of 28	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.