Qwen 3 235B
publishedAlibaba's flagship reasoning model
Large MoE model with strong math reasoning. 92.3% on AIME 2025. Thinking mode with Apache 2.0 license.
Provider alibaba Type reasoning Access open_source Params 235B MoE Context 128k License apache-2.0
Benchmarks (1)
| Benchmark | Score | Reported By | Source |
|---|---|---|---|
| AIME 2025 | 92.3% | vendor | source ↗ |
Why It Matters
Highest AIME 2025 score among open-source models. Apache 2.0 — truly open with no commercial restrictions.
Known Limitations
Large model requires significant GPU resources. Chinese regulatory environment.
Provider alibaba
Released 2025-05-01
Training cutoff 2025-04
Created March 22, 2026
Last reconciled Never