Compare / head-to-head
Qwen3 235B-A22B Instruct 2507vsQwen3 235B-A22B Thinking 2507
Qwen3 235B-A22B Thinking 2507 leads 11 of 14 shared benchmarks. Qwen3 235B-A22B Instruct 2507 is 2.2x cheaper per token. Both offer a 262K-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
3 – 11
Qwen3 235B-A22B Thinking 2507 leads
Cheaper per token
Qwen3 235B-A22B Instruct 2507
2.2x cheaper, input + output
Larger context
Tie
both 262K tokens
Providers
Qwen
same provider
Qwen3 235B-A22B Instruct 2507
Qwen · released 2025-07-21
text
Qwen3 235B-A22B Thinking 2507
Qwen · released 2025-07-25
text
Quality
Benchmark matrix
14 shared · 3 only Qwen3 235B-A22B Instruct 2507 · 9 only Qwen3 235B-A22B Thinking 2507| Benchmark | Qwen3 235B-A22B Instruct 2507 | Qwen3 235B-A22B Thinking 2507 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 14 | ||||
| MMLU-Pro | 83% ↗ | 84.4% ↗ | -1.4 pt | Qwen3 235B-A22B Thinking 2507 |
| LMArena Elo | 1422.4 ↗ | 1400 ↗ | +22.4 | Qwen3 235B-A22B Instruct 2507 |
| AIME25 | 70.3% ↗ | 92.3% ↗ | -22 pt | Qwen3 235B-A22B Thinking 2507 |
| Arena-Hard v2 | 79.2% ↗ | 79.7% ↗ | -0.5 pt | Qwen3 235B-A22B Thinking 2507 |
| BFCL-v3 | 70.9% ↗ | 71.9% ↗ | -1 pt | Qwen3 235B-A22B Thinking 2507 |
| GPQA | 77.5% ↗ | 81.1% ↗ | -3.6 pt | Qwen3 235B-A22B Thinking 2507 |
| HMMT25 | 55.4% ↗ | 83.9% ↗ | -28.5 pt | Qwen3 235B-A22B Thinking 2507 |
| IFEval | 88.7% ↗ | 87.8% ↗ | +0.9 pt | Qwen3 235B-A22B Instruct 2507 |
| LiveBench 20241125 | 75.4% ↗ | 78.4% ↗ | -3 pt | Qwen3 235B-A22B Thinking 2507 |
| LiveCodeBench v6 (25.02-25.05) | 51.8% ↗ | 74.1% ↗ | -22.3 pt | Qwen3 235B-A22B Thinking 2507 |
| MMLU-Redux | 93.1% ↗ | 93.8% ↗ | -0.7 pt | Qwen3 235B-A22B Thinking 2507 |
| SuperGPQA | 62.6% ↗ | 64.9% ↗ | -2.3 pt | Qwen3 235B-A22B Thinking 2507 |
| TAU2-Retail | 74.6% ↗ | 71.9% ↗ | +2.7 pt | Qwen3 235B-A22B Instruct 2507 |
| WeirdML (Håvard Tveit Ihle) | 38.7% ↗ | 41% ↗ | -2.3 pt | Qwen3 235B-A22B Thinking 2507 |
| Only Qwen3 235B-A22B Instruct 2507 reports · 3 | ||||
| Aider-Polyglot | 57.3% ↗ | not reported | — | — |
| ARC-AGI | 41.8% ↗ | not reported | — | — |
| SimpleQA | 54.3% ↗ | not reported | — | — |
| Only Qwen3 235B-A22B Thinking 2507 reports · 9 | ||||
| GPQA Diamond | not reported | 80.1% ↗epoch run | — | — |
| Chess Puzzles (Epoch AI run) | not reported | 12% ↗epoch run | — | — |
| HLE (text-only) | not reported | 18.2% ↗ | — | — |
| Mystery Game Puzzles (Epoch AI run) | not reported | 9% ↗epoch run | — | — |
| OJBench | not reported | 32.5% ↗ | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | not reported | 86.7% ↗epoch run | — | — |
| SimpleQA Verified | not reported | 40.4% ↗epoch run | — | — |
| TAU2-Airline | not reported | 58% ↗ | — | — |
| Vending-Bench 2 (Andon Labs) | not reported | -11.34 ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Qwen3 235B-A22B Instruct 2507 minus Qwen3 235B-A22B Thinking 2507 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Qwen3 235B-A22B Instruct 2507 | Qwen3 235B-A22B Thinking 2507 | Edge |
|---|---|---|---|
| Context window | 262K tokens | 262K tokens | Tie |
| Max output | — | — | — |
| Input price / 1M | $0.23 ↗ | $0.23 ↗ | Tie |
| Output price / 1M | $0.92 ↗ | $2.3 ↗ | Qwen3 235B-A22B Instruct 2507 |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $1.15 | $2.53 | Qwen3 235B-A22B Instruct 2507 |
| Modalities | text | text | Tie |
| Released | 2025-07-21 | 2025-07-25 | — |
| Cited benchmark scores | 17 | 23 | — |
Reliability
Qwen status
Live from /statusMore matchups
Qwen3 235B-A22B Instruct 2507 vs …
Models sharing the most benchmarksMore matchups
Qwen3 235B-A22B Thinking 2507 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Qwen3 235B-A22B Instruct 2507 and Qwen3 235B-A22B Thinking 2507 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.