Compare / head-to-head
Qwen3.5 0.8BvsQwen3.5 2B
Qwen3.5 2B leads 14 of 14 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
0 – 14
Qwen3.5 2B leads
Cheaper per token
—
list price, input + output
Larger context
Tie
both 262K tokens
Providers
Qwen
same provider
Qwen3.5 0.8B
Qwen · released 2026-03-02
textvisionvideo
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 14 · 5 core
Qwen3.5 2B
Qwen · released 2026-03-02
textvisionvideo
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 16 · 5 core
Quality
Benchmark matrix
14 shared · 0 only Qwen3.5 0.8B · 2 only Qwen3.5 2B| Benchmark | Qwen3.5 0.8B | Qwen3.5 2B | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 14 | ||||
| BFCL-V4 (thinking) | 25.3% ↗ | 43.6% ↗ | -18.3 pt | Qwen3.5 2B |
| GPQA (thinking) | 11.9% ↗ | 51.6% ↗ | -39.7 pt | Qwen3.5 2B |
| IFBench (thinking) | 21% ↗ | 41.3% ↗ | -20.3 pt | Qwen3.5 2B |
| IFEval (non-thinking) | 52.1% ↗ | 61.2% ↗ | -9.1 pt | Qwen3.5 2B |
| IFEval (thinking) | 44% ↗ | 78.6% ↗ | -34.6 pt | Qwen3.5 2B |
| LongBench v2 (thinking) | 26.1% ↗ | 38.7% ↗ | -12.6 pt | Qwen3.5 2B |
| MMLU-Pro (non-thinking) | 29.7% ↗ | 55.3% ↗ | -25.6 pt | Qwen3.5 2B |
| MMLU-Pro (thinking) | 42.3% ↗ | 66.5% ↗ | -24.2 pt | Qwen3.5 2B |
| MMLU-Redux (thinking) | 59.5% ↗ | 79.6% ↗ | -20.1 pt | Qwen3.5 2B |
| MMMLU (thinking) | 44.3% ↗ | 63.1% ↗ | -18.8 pt | Qwen3.5 2B |
| MMMU (thinking) | 49% ↗ | 64.2% ↗ | -15.2 pt | Qwen3.5 2B |
| MMMU-Pro (thinking) | 31.2% ↗ | 50.3% ↗ | -19.1 pt | Qwen3.5 2B |
| SuperGPQA (thinking) | 21.3% ↗ | 37.5% ↗ | -16.2 pt | Qwen3.5 2B |
| TAU2-Bench (thinking) | 11.6% ↗ | 48.8% ↗ | -37.2 pt | Qwen3.5 2B |
| Only Qwen3.5 0.8B reports · 0 | ||||
| Qwen3.5 0.8B reports nothing Qwen3.5 2B does not. | ||||
| Only Qwen3.5 2B reports · 2 | ||||
| Chess Puzzles (Epoch AI run) | not reported | 0% ↗epoch run | — | — |
| HMMT Feb 25 (thinking) | not reported | 22.9% ↗ | — | — |
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 0.8B minus Qwen3.5 2B 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 0.8B | Qwen3.5 2B | 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 | 14 | 16 | — |
Reliability
Qwen status
Live from /statusMore matchups
Qwen3.5 0.8B vs …
Models sharing the most benchmarksMore matchups
Qwen3.5 2B vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Qwen3.5 0.8B and Qwen3.5 2B on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.