Qwen·released 2025-09-241 source
Qwen3-Max benchmark scores: 11 benchmarks tracked. 36% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 11 benchmarks·best result 97.13 on MATH level 5 (reproduced)·4 of 11 independently reproduced·$2.11/$8.45 per M tokens
Consensus: LiteLLM · Cross-check: models.dev · See every model’s pricing →
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| MATH level 5 | 97.13accuracy (%) | reproduced· optimizedT1 | 2025-09-24Epoch AI |
| OTIS Mock AIME 2024-2025 | 73.31accuracy (%) | reproduced· optimizedT1 | 2025-09-24Epoch AI |
| DTBench | 70.22accuracy (%) | unverifiedT1 | 2025-09-24https://conceptualreasoning.ai/dtbench |
| Fiction.LiveBench | 66.7accuracy (%) | unverifiedT1 | 2025-09-24Fiction.live leaderboard |
| GPQA diamond | 63.47accuracy (%) | reproduced· optimizedT1 | 2025-09-24Epoch AI |
| SimpleQA Verified | 48.75accuracy (%) | unverifiedT2 | 2025-09-24 |
| LMCA | 33.26accuracy (%) | unverifiedT1 | 2025-09-24https://conceptualreasoning.ai/lmca |
| FrontierMath-Tiers-1-3-v2-Private | 18.95accuracy (%) | unverifiedT2 | 2025-09-24 |
| CL-bench | 14.5accuracy (%) | unverifiedT2 | 2025-09-24 |
| Chess Puzzles | 0accuracy (%) | reproducedT1 | 2025-09-24Epoch AI |
| Mystery Game Puzzles | 0accuracy (%) | unverifiedT2 | 2025-09-24 |
Claims drawn from cited facts, not live model generation.
This model was developed by Qwen and released in 2025, placing it among the current generation of offerings rather than an older line. No parameter count is disclosed, so its scale cannot be quantified here; its recency is the defining characteristic of its vintage.
3 cited facts
This model is scored on 11 tracked benchmarks, and it currently holds no top score on any of them. Its record can be trusted: the scores have been independently reproduced, meaning outside evaluation backs them up.
3 cited facts
The model trails the benchmark leader by an average of 38.66 points under the disclosed harness — a wide gap that reads as well behind the leaders. It holds the top score on 0 benchmarks, meaning it shows no category leadership yet.
2 cited facts
6 facts cross-checked across data sources: 2 single-source, 4 disputed.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks, litellm_prices, modelsdev_models
Qwen3-Max is an AI model developed by Qwen, released 2025-09-24. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Qwen3-Max has recorded scores on 11 benchmarks, each shown with its evidence status.
Qwen3-Max has recorded scores on 11 benchmarks — GPQA diamond, MATH level 5, OTIS Mock AIME 2024-2025, DTBench, LMCA, SimpleQA Verified, and 5 more. The full table above shows each score with its evidence status.
4 of 11 recorded scores (36%) are independently reproduced rather than self-reported by the lab.
Qwen3-Max has 11 tracked claims: 4 independently reproduced, 7 unverified.
Listed API pricing: $2.11 per million input tokens, $8.45 per million output tokens (prices disputed across sources). See the pricing block for the full breakdown.