Alibaba
The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall...
Qwen model releasesRank #189 across 566
No rank
Rank #135 across 371
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Higher bars mean stronger relative placement.
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 23.4 | #189 |
| GPQA | reasoning | 81.9% | #127 |
| Humanity's Last Exam | reasoning | 12.8% | #160 |
| SciCode | coding, reasoning | 29.3% | #311 |
| Output Speed |
| speed |
| 166.4 tok/s |
| #60 |
| Time to First Token | speed | 1.17s | #176 |
| Blended Price | cost | $0.688/M | #135 |
| Input Price | cost | $0.250/M | #104 |
| Output Price | cost | $2.00/M | #153 |
| Value Index | cost, overall | 34.0 | #106 |