Compare / head-to-head
Qwen3.5 4BvsQwen3.5 9B
Qwen3.5 9B leads 15 of 16 shared benchmarks. Both offer a 262K-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
1 – 15
Qwen3.5 9B leads
Cheaper per token
—
list price, input + output
Larger context
Tie
both 262K tokens
Providers
Qwen
same provider
Qwen3.5 4B
Qwen · released 2026-03-02
textvisionvideo
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 16 · 5 core
Qwen3.5 9B
Qwen · released 2026-03-02
textvisionvideo
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 18 · 5 core
Quality
Benchmark matrix
16 shared · 0 only Qwen3.5 4B · 1 only Qwen3.5 9B| Benchmark | Qwen3.5 4B | Qwen3.5 9B | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 16 | ||||
| GPQA Diamond | 76.2% ↗ | 81.7% ↗ | -5.5 pt | Qwen3.5 9B |
| MMLU-Pro | 79.1% ↗ | 82.5% ↗ | -3.4 pt | Qwen3.5 9B |
| BFCL-V4 | 50.3% ↗ | 66.1% ↗ | -15.8 pt | Qwen3.5 9B |
| HMMT Feb 25 | 74% ↗ | 83.2% ↗ | -9.2 pt | Qwen3.5 9B |
| IFBench | 59.2% ↗ | 64.5% ↗ | -5.3 pt | Qwen3.5 9B |
| IFEval | 89.8% ↗ | 91.5% ↗ | -1.7 pt | Qwen3.5 9B |
| LiveCodeBench v6 | 55.8% ↗ | 65.6% ↗ | -9.8 pt | Qwen3.5 9B |
| LongBench v2 | 50% ↗ | 55.2% ↗ | -5.2 pt | Qwen3.5 9B |
| MMLU-Redux | 88.8% ↗ | 91.1% ↗ | -2.3 pt | Qwen3.5 9B |
| MMMLU | 76.1% ↗ | 81.2% ↗ | -5.1 pt | Qwen3.5 9B |
| MMMU | 77.6% ↗ | 78.4% ↗ | -0.8 pt | Qwen3.5 9B |
| MMMU-Pro | 66.3% ↗ | 70.1% ↗ | -3.8 pt | Qwen3.5 9B |
| OSWorld-Verified | 35.6% ↗ | 41.8% ↗ | -6.2 pt | Qwen3.5 9B |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 55.8% ↗epoch run | 61.7% ↗epoch run | -5.9 pt | Qwen3.5 9B |
| SuperGPQA | 52.9% ↗ | 58.2% ↗ | -5.3 pt | Qwen3.5 9B |
| TAU2-Bench | 79.9% ↗ | 79.1% ↗ | +0.8 pt | Qwen3.5 4B |
| Only Qwen3.5 4B reports · 0 | ||||
| Qwen3.5 4B reports nothing Qwen3.5 9B does not. | ||||
| Only Qwen3.5 9B reports · 1 | ||||
| Chess Puzzles (Epoch AI run) | not reported | 12% ↗epoch run | — | — |
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 Qwen3.5 4B minus Qwen3.5 9B 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 | Qwen3.5 4B | Qwen3.5 9B | Edge |
|---|---|---|---|
| Context window | 262K tokens | 262K tokens | Tie |
| Max output | — | — | — |
| Input price / 1M | — | — | — |
| Output price / 1M | — | — | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | — | — |
| Modalities | text · vision · video | text · vision · video | Tie |
| Released | 2026-03-02 | 2026-03-02 | — |
| Cited benchmark scores | 16 | 18 | — |
Reliability
Qwen status
Live from /statusMore matchups
Qwen3.5 4B vs …
Models sharing the most benchmarksMore matchups
Qwen3.5 9B vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Qwen3.5 4B and Qwen3.5 9B on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.