Qwen·released 2025-07-251 source
Qwen3-235B-A22B-Thinking (Jul 2025) benchmark scores: 12 benchmarks tracked. 42% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 12 benchmarks·best result 86.65 on OTIS Mock AIME 2024-2025 (reproduced)·5 of 12 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.65accuracy (%) | reproduced· optimizedT1 | 2025-07-25Epoch AI |
| Lech Mazur Writing | 82.4accuracy (%) | unverifiedT1 | 2025-07-25lechmazur/writing Github repository |
| Fiction.LiveBench | 75accuracy (%) | unverifiedT1 | 2025-07-25Fiction.live leaderboard |
| GPQA diamond | 73.4accuracy (%) | reproduced· optimizedT1 | 2025-07-25Epoch AI |
| DTBench | 67.12accuracy (%) | unverifiedT1 | 2025-07-25https://conceptualreasoning.ai/dtbench |
| WeirdML | 41.04accuracy (%) | unverifiedT1 | 2025-07-25WeirdML Leaderboard |
| SimpleQA Verified | 40.44accuracy (%) | unverifiedT2 | 2025-07-25 |
| LMCA | 34.48accuracy (%) | unverifiedT1 | 2025-07-25https://conceptualreasoning.ai/lmca |
| FrontierMath-2025-02-28-Private | 14.88accuracy (%) | reproduced· optimizedT1 | 2025-07-25Epoch AI |
| Chess Puzzles | 7.41accuracy (%) | reproducedT1 | 2025-07-25Epoch AI |
| FrontierMath-Tier-4-2025-07-01-Private | 0accuracy (%) | reproducedT1 | 2025-07-25Epoch AI |
| Mystery Game Puzzles | 0accuracy (%) | unverifiedT2 | 2025-07-25 |
Claims drawn from cited facts, not live model generation.
Developed by Qwen, this model is a large, frontier-scale 235-billion-parameter system released in 2025. That scale implies more capability headroom but also greater inference and deployment expense than compact alternatives, trading efficiency and lower cost for higher cost at the large end.
4 cited facts
This model is scored on 12 tracked benchmarks, but it currently holds no top score on any of them. Because its results have been independently reproduced, outside evaluation backs up those scores.
3 cited facts
On average this model trails the benchmark leader by 37.3 points, a wide gap that reads as well behind the front of the pack rather than competitive. It holds the top score on none of its benchmarks, so there is no category leadership to point to here.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: curated_param_counts, epoch_benchmarks
Qwen3-235B-A22B-Thinking (Jul 2025) is an AI model developed by Qwen, released 2025-07-25. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Qwen3-235B-A22B-Thinking (Jul 2025) has recorded scores on 12 benchmarks, each shown with its evidence status.
Qwen3-235B-A22B-Thinking (Jul 2025) has recorded scores on 12 benchmarks — Lech Mazur Writing, GPQA diamond, OTIS Mock AIME 2024-2025, DTBench, WeirdML, LMCA, and 6 more. The full table above shows each score with its evidence status.
5 of 12 recorded scores (42%) are independently reproduced rather than self-reported by the lab.
Qwen3-235B-A22B-Thinking (Jul 2025) has 12 tracked claims: 5 independently reproduced, 7 unverified.