Model comparison
Qwen3.5-9B vs Qwen3.5-Flash
Qwen3.5-Flash is the stronger model overall, scoring 42.3 to 33.3 on the Noometry Index. Qwen3.5-9B costs 1.6× less per token, which makes it the better buy when Qwen3.5-Flash's lead doesn't matter for your workload.
Last verified . 6 shared benchmarks.
Alibaba (Qwen)Summary
- They share 6 benchmarks with published results for both. Qwen3.5-9B scores higher in 2 categories and Qwen3.5-Flash in 2 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5-Flash leads 32.9 to 22.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 61.7% for Qwen3.5-9B and 84.4% for Qwen3.5-Flash.
- Qwen3.5-9B is cheaper at $0.10 / $0.15 per million input/output tokens, against $0.10 / $0.40 for Qwen3.5-Flash.
- Qwen3.5-Flash accepts more context: 1M tokens versus 262K.
- Qwen3.5-9B has downloadable open weights; the other is API-only.
Side by side
| Qwen3.5-9B | Qwen3.5-Flash | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 33.3 | 42.3 |
| Released | 2026-02-23 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 262K | 1M |
| Max output | 66K | 66K |
| Input $ / M tokens | $0.10 | $0.10 |
| Output $ / M tokens | $0.15 | $0.40 |
| Results tracked | 10 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5-9B leads
Qwen3.5-9B: 35.3 (#213), Qwen3.5-Flash: 33.9 (#234)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| LMArena WebDev | — | 1243 |
| SciCode | 27.5% | — |
| LMArena Coding | — | 1412 |
| ALE-Bench | — | 221.8 |
Agentic & Tool Use Not comparable
Qwen3.5-9B: 14.4 (#143), Qwen3.5-Flash: —
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| Terminal-Bench | 9.2% | — |
| Vending-Bench 2 | — | 462.69 |
Reasoning Qwen3.5-Flash leads
Qwen3.5-9B: 22.5 (#178), Qwen3.5-Flash: 32.9 (#63)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| Chess Puzzles | 12% | 21% |
| DTBench | 71.2% | 82.9% |
| LMCA | 24.5% | 29.1% |
| Epoch Capabilities Index | 139.45 | 143.98 |
| CritPt | 0.3% | — |
| LMArena Hard Prompts | — | 1404 |
| Mystery Game Puzzles | — | 20% |
Math Qwen3.5-Flash leads
Qwen3.5-9B: 33.5 (#174), Qwen3.5-Flash: 36.3 (#117)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 61.7% | 84.4% |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| MathArena Final-Answer Competitions | 48.5% | — |
| LMArena Math | — | 1407 |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-9B leads
Qwen3.5-9B: 46.0 (#77), Qwen3.5-Flash: 43.2 (#84)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 79% | 82.3% |
| SimpleQA Verified | — | 20.3% |
| Vectara Hallucination Rate | — | 10.5% |
| LMArena Expert | — | 1408 |
Multilingual Not comparable
Qwen3.5-9B: —, Qwen3.5-Flash: 50.7 (#116)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | — | 1385 |
| LMArena Chinese | — | 1446 |
| LMArena French | — | 1412 |
| LMArena German | — | 1389 |
| LMArena Japanese | — | 1369 |
| LMArena Korean | — | 1344 |
| LMArena Russian | — | 1379 |
| LMArena Spanish | — | 1399 |
Instruction Following Not comparable
Qwen3.5-9B: —, Qwen3.5-Flash: 72.8 (#133)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | — | 1374 |
Long Context Not comparable
Qwen3.5-9B: —, Qwen3.5-Flash: 42.4 (#119)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | — | 1392 |
Writing & Preference Not comparable
Qwen3.5-9B: —, Qwen3.5-Flash: 58.2 (#117)
| Benchmark | Qwen3.5-9B | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | — | 1397 |
| LMArena Creative Writing | — | 1343 |
| LMArena Multi-Turn | — | 1393 |
Frequently asked questions
Is Qwen3.5-9B better than Qwen3.5-Flash?
Qwen3.5-Flash is the stronger model overall, scoring 42.3 to 33.3 on the Noometry Index. Qwen3.5-9B costs 1.6× less per token, which makes it the better buy when Qwen3.5-Flash's lead doesn't matter for your workload.
Which is cheaper, Qwen3.5-9B or Qwen3.5-Flash?
Qwen3.5-9B is cheaper. It lists at $0.10 per million input tokens and $0.15 per million output tokens; Qwen3.5-Flash lists at $0.10 and $0.40.
Is Qwen3.5-9B or Qwen3.5-Flash better for coding?
Qwen3.5-9B scores higher on coding benchmarks: 35.3 versus 33.9 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5-Flash does, with 1M tokens against 262K.
How many benchmarks do Qwen3.5-9B and Qwen3.5-Flash share?
6 benchmarks have published results for both models. Qwen3.5-9B has 10 scored results on Noometry and Qwen3.5-Flash has 32.