Compare / head-to-head
Qwen3-VL 235B-A22B InstructvsQwen3-VL 235B-A22B Thinking
Qwen3-VL 235B-A22B Thinking leads 9 of 13 shared benchmarks. Qwen3-VL 235B-A22B Instruct is 2.2x cheaper per token. Both offer a 256K-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
4 – 9
Qwen3-VL 235B-A22B Thinking leads
Cheaper per token
Qwen3-VL 235B-A22B Instruct
2.2x cheaper, input + output
Larger context
Tie
both 256K tokens
Providers
Qwen
same provider
Qwen3-VL 235B-A22B Instruct
Qwen · released 2025-09-23
textvisionvideo
Qwen3-VL 235B-A22B Thinking
Qwen · released 2025-09-23
textvisionvideo
Quality
Benchmark matrix
13 shared · 4 only Qwen3-VL 235B-A22B Instruct · 4 only Qwen3-VL 235B-A22B Thinking| Benchmark | Qwen3-VL 235B-A22B Instruct | Qwen3-VL 235B-A22B Thinking | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 13 | ||||
| MMLU-Pro | 81.8% ↗ | 83.8% ↗ | -2 pt | Qwen3-VL 235B-A22B Thinking |
| LMArena Elo | 1413.4 ↗ | 1395 ↗ | +18.4 | Qwen3-VL 235B-A22B Instruct |
| AIME25 | 74.7% ↗ | 89.7% ↗ | -15 pt | Qwen3-VL 235B-A22B Thinking |
| HMMT25 | 57.4% ↗ | 77.4% ↗ | -20 pt | Qwen3-VL 235B-A22B Thinking |
| IFEval | 87.8% ↗ | 88.2% ↗ | -0.4 pt | Qwen3-VL 235B-A22B Thinking |
| LMArena Vision (LMArena) | 1215.1 ↗ | 1188 ↗ | +27.1 | Qwen3-VL 235B-A22B Instruct |
| MathVista (mini) | 84.9% ↗ | 85.8% ↗ | -0.9 pt | Qwen3-VL 235B-A22B Thinking |
| MMLU-Redux | 92.2% ↗ | 93.7% ↗ | -1.5 pt | Qwen3-VL 235B-A22B Thinking |
| MMMU (val) | 78.7% ↗ | 80.6% ↗ | -1.9 pt | Qwen3-VL 235B-A22B Thinking |
| MMMU_Pro | 68.1% ↗ | 69.3% ↗ | -1.2 pt | Qwen3-VL 235B-A22B Thinking |
| SimpleQA | 51.9% ↗ | 44.4% ↗ | +7.5 pt | Qwen3-VL 235B-A22B Instruct |
| SuperGPQA | 60.4% ↗ | 64.3% ↗ | -3.9 pt | Qwen3-VL 235B-A22B Thinking |
| VideoMME (w/o sub) | 79.2% ↗ | 79% ↗ | +0.2 pt | Qwen3-VL 235B-A22B Instruct |
| Only Qwen3-VL 235B-A22B Instruct reports · 4 | ||||
| Arena-Hard v2 | 77.4% ↗ | not reported | — | — |
| BFCL-v3 | 67.7% ↗ | not reported | — | — |
| LCBV6 (25.02-25.05) | 54.3% ↗ | not reported | — | — |
| ScreenSpot Pro | 62% ↗ | not reported | — | — |
| Only Qwen3-VL 235B-A22B Thinking reports · 4 | ||||
| BFCL | not reported | 71.9% ↗ | — | — |
| HLE | not reported | 13.6% ↗ | — | — |
| LCB v6 | not reported | 70.1% ↗ | — | — |
| OSWorld | not reported | 38.1% ↗ | — | — |
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-VL 235B-A22B Instruct minus Qwen3-VL 235B-A22B Thinking 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-VL 235B-A22B Instruct | Qwen3-VL 235B-A22B Thinking | Edge |
|---|---|---|---|
| Context window | 256K tokens | 256K tokens | Tie |
| Max output | — | — | — |
| Input price / 1M | $0.4 ↗ | $0.4 ↗ | Tie |
| Output price / 1M | $1.6 ↗ | $4 ↗ | Qwen3-VL 235B-A22B Instruct |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $2 | $4.4 | Qwen3-VL 235B-A22B Instruct |
| Modalities | text · vision · video | text · vision · video | Tie |
| Released | 2025-09-23 | 2025-09-23 | — |
| Cited benchmark scores | 17 | 17 | — |
Reliability
Qwen status
Live from /statusMore matchups
Qwen3-VL 235B-A22B Instruct vs …
Models sharing the most benchmarksMore matchups
Qwen3-VL 235B-A22B Thinking vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Qwen3-VL 235B-A22B Instruct and Qwen3-VL 235B-A22B Thinking on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.