Qwen3 235B-A22B Thinking 2507 vs Qwen3.5 35B-A3B
| Spec | Qwen3 235B-A22B Thinking 2507 | Qwen3.5 35B-A3B |
|---|---|---|
| Respan Index | 18.5 (#142) | 18.4 (#144) |
| Blended price per 1M | $1.21 | $0.158 |
| Context window | 262K | 262K |
| Input / output per 1M | $0.45 / $3.5 | $0.057 / $0.459 |
| Max output | - | - |
| Input | text | text, vision |
| Released | 2025-07-25 | - |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Qwen3 235B-A22B Thinking 2507 | Qwen3.5 35B-A3B |
|---|---|---|
| Coding | 5.2 Qwen3 235B-A22B Thinking 2507 | 0.9 Qwen3.5 35B-A3B |
| Agents | 4.1 Qwen3 235B-A22B Thinking 2507 | 15.5* Qwen3.5 35B-A3B |
| Reasoning | 22.1 Qwen3 235B-A22B Thinking 2507 | 23.4 Qwen3.5 35B-A3B |
| Knowledge & language | 38.1 Qwen3 235B-A22B Thinking 2507 | 13.9* Qwen3.5 35B-A3B |
| Human preference | 45.4 Qwen3 235B-A22B Thinking 2507 | 47.1 Qwen3.5 35B-A3B |
* Estimated: no published scores in that category. How the Respan Index works
Benchmarks both report
Qwen3 235B-A22B Thinking 2507 leads on 3 of the 4 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Qwen3 235B-A22B Thinking 2507 | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Elo | 1400 | 1394 |
| Chess Puzzles (Epoch AI) | 12% | 10% |
| GPQA Diamond | 80.1% | 83.5% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 86.7% | 70% |
FAQ
Is Qwen3 235B-A22B Thinking 2507 better than Qwen3.5 35B-A3B?
Qwen3 235B-A22B Thinking 2507 scores 18.5 on the Respan Index (#142), ahead of Qwen3.5 35B-A3B at 18.4 (#144). Qwen3 235B-A22B Thinking 2507 leads on 3 of the 4 benchmarks both report.
Which is cheaper, Qwen3 235B-A22B Thinking 2507 or Qwen3.5 35B-A3B?
Qwen3 235B-A22B Thinking 2507 lists at $0.45 input and $3.5 output per 1M tokens; Qwen3.5 35B-A3B at $0.057 and $0.459. Qwen3.5 35B-A3B is 7.7x cheaper per token at list price.
Which has the larger context window?
Both accept 262K tokens of context.