Qwen 3.5 Flash (hosted 35B-A3B)
Qwen 3.5 Flash (hosted 35B-A3B).
Qwen Released 25 Feb 2026 26
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Qwen Feb 2026 Qwen 3.5 Flash (hosted 35B-A3B)
143.6
#103 of 271
Epoch Capabilities Index
- GPQA Diamond 82.3% #83 /219
- FrontierMath 6.21% #42 /70
- FrontierMath Tier 4 0% #54 /55
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.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vectara Hallucination Leaderboard (2025 dataset) | 10.5% | #65 of 104 | Official leaderboard | |
| SimpleQA Verified | 20.3% | #71 of 84 | Epoch AI |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 82.3% | #83 of 219 | Epoch AI | |
| OTIS Mock AIME 2024-2025 | 84.4% | #80 of 202 | Epoch AI | |
| DTBench | 71.5% | #75 of 178 | Official leaderboard | |
| FrontierMath | 6.21% | #42 of 70 | Epoch AI | |
| FrontierMath Tiers 1-3 v2 | 18.3% | #81 of 89 | Epoch AI | |
| FrontierMath Tier 4 | 0% | #54 of 55 | Epoch AI |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| WebDev Arena | 1,238 | #97 of 105 | Epoch AI | |
| ALE-Bench | 222 | #117 of 121 | Epoch AI |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Mystery Game Puzzles | 20% | #45 of 80 | Epoch AI | |
| Chess Puzzles | 14% | #69 of 83 | Epoch AI |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LegalBench | 84.3% | #36 of 138 | Official leaderboard | |
| MortgageTax | 67.4% | #24 of 86 | Official leaderboard | |
| CorpFin v2 | 63.6% | #45 of 119 | Official leaderboard | |
| TaxEval v2 | 72.2% | #64 of 129 | Official leaderboard | |
| CaseLaw v2 | 56% | #31 of 53 | Official leaderboard | |
| Vals Finance Agent | 45.6% | #37 of 49 | Official leaderboard | |
| MedCode | 33% | #79 of 96 | Official leaderboard | |
| Vending-Bench 2 | $463 | #55 of 65 | Epoch AI | |
| MedScribe | 70.6% | #85 of 98 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Conceptual Reasoning Index | 42.3 | #91 of 149 | Official leaderboard | |
| LMCA | 29.1% | #85 of 139 | Epoch AI | |
| SAGE v1.1 (Vals) | 84.9% | #18 of 28 | Official leaderboard | |
| SAGE (Vals) | 42.5% | #53 of 81 | Official leaderboard | |
| ACCoRD | 52.5 | #109 of 150 | Official leaderboard |
Qwen's models over time
Qwen's models in color, against every other model on the Epoch Capabilities Index, by release date.