Compare / head-to-head
Qwen3 30B-A3B Instruct 2507vsQwen3 4B Thinking 2507
Qwen3 30B-A3B Instruct 2507 and Qwen3 4B Thinking 2507 split their 12 shared benchmarks evenly. Both offer a 262K-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
6 – 6
even split
Cheaper per token
—
list price, input + output
Larger context
Tie
both 262K tokens
Providers
Qwen
same provider
Qwen3 30B-A3B Instruct 2507
Qwen · released 2025-07-30
text
Qwen3 4B Thinking 2507
Qwen · released 2025-08-06
text
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 15 · 5 core
Quality
Benchmark matrix
12 shared · 7 only Qwen3 30B-A3B Instruct 2507 · 3 only Qwen3 4B Thinking 2507| Benchmark | Qwen3 30B-A3B Instruct 2507 | Qwen3 4B Thinking 2507 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 12 | ||||
| MMLU-Pro | 78.4% ↗ | 74% ↗ | +4.4 pt | Qwen3 30B-A3B Instruct 2507 |
| AIME25 | 61.3% ↗ | 81.3% ↗ | -20 pt | Qwen3 4B Thinking 2507 |
| Arena-Hard v2 | 69% ↗ | 34.9% ↗ | +34.1 pt | Qwen3 30B-A3B Instruct 2507 |
| BFCL-v3 | 65.1% ↗ | 71.2% ↗ | -6.1 pt | Qwen3 4B Thinking 2507 |
| GPQA | 70.4% ↗ | 65.8% ↗ | +4.6 pt | Qwen3 30B-A3B Instruct 2507 |
| HMMT25 | 43% ↗ | 55.5% ↗ | -12.5 pt | Qwen3 4B Thinking 2507 |
| IFEval | 84.7% ↗ | 87.4% ↗ | -2.7 pt | Qwen3 4B Thinking 2507 |
| LiveBench 20241125 | 69% ↗ | 71.8% ↗ | -2.8 pt | Qwen3 4B Thinking 2507 |
| LiveCodeBench v6 (25.02-25.05) | 43.2% ↗ | 55.2% ↗ | -12 pt | Qwen3 4B Thinking 2507 |
| MMLU-Redux | 89.3% ↗ | 86.1% ↗ | +3.2 pt | Qwen3 30B-A3B Instruct 2507 |
| SuperGPQA | 53.4% ↗ | 47.8% ↗ | +5.6 pt | Qwen3 30B-A3B Instruct 2507 |
| TAU2-Retail | 57% ↗ | 53.5% ↗ | +3.5 pt | Qwen3 30B-A3B Instruct 2507 |
| Only Qwen3 30B-A3B Instruct 2507 reports · 7 | ||||
| GPQA Diamond | 55.6% ↗epoch run | not reported | — | — |
| LMArena Elo | 1382.7 ↗ | not reported | — | — |
| Aider-Polyglot | 35.6% ↗ | not reported | — | — |
| Chess Puzzles (Epoch AI run) | 2% ↗epoch run | not reported | — | — |
| MultiPL-E | 83.8% ↗ | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 62.2% ↗epoch run | not reported | — | — |
| ZebraLogic | 90% ↗ | not reported | — | — |
| Only Qwen3 4B Thinking 2507 reports · 3 | ||||
| OJBench | not reported | 17.9% ↗ | — | — |
| TAU2-Airline | not reported | 58% ↗ | — | — |
| TAU2-Telecom | not reported | 27.2% ↗ | — | — |
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 30B-A3B Instruct 2507 minus Qwen3 4B 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 30B-A3B Instruct 2507 | Qwen3 4B Thinking 2507 | Edge |
|---|---|---|---|
| Context window | 262K tokens | 262K tokens | Tie |
| Max output | — | — | — |
| Input price / 1M | $0.2 ↗ | — | — |
| Output price / 1M | $0.8 ↗ | — | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $1 | — | — |
| Modalities | text | text | Tie |
| Released | 2025-07-30 | 2025-08-06 | — |
| Cited benchmark scores | 19 | 15 | — |
Reliability
Qwen status
Live from /statusMore matchups
Qwen3 30B-A3B Instruct 2507 vs …
Models sharing the most benchmarksMore matchups
Qwen3 4B Thinking 2507 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Qwen3 30B-A3B Instruct 2507 and Qwen3 4B Thinking 2507 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.