Model comparison
Qwen3-4B vs Qwen3.5-Flash
Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 31.9 on the Noometry Index.
Last verified . 4 shared benchmarks.
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. Qwen3-4B scores higher in 0 categories and Qwen3.5-Flash in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5-Flash leads 33.7 to 19.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 52.2% for Qwen3-4B and 84.4% for Qwen3.5-Flash.
- Qwen3-4B has downloadable open weights; the other is API-only.
Side by side
| Qwen3-4B | Qwen3.5-Flash | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 42.5 |
| Released | 2025-04-29 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | — | 1M |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.40 |
| Results tracked | 6 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Qwen3-4B: —, Qwen3.5-Flash: 34.2 (#242)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| LMArena WebDev | — | 1244 |
| LMArena Coding | — | 1412 |
| ALE-Bench | — | 221.8 |
Agentic & Tool Use Not comparable
Qwen3-4B: 27.6 (#100), Qwen3.5-Flash: —
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| Berkeley Function Calling Leaderboard | 35.7% | — |
| Vending-Bench 2 | — | 462.69 |
Reasoning Qwen3.5-Flash leads
Qwen3-4B: 19.2 (#268), Qwen3.5-Flash: 33.7 (#72)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| Chess Puzzles | 4% | 21% |
| LMArena Hard Prompts | — | 1403 |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 82.9% |
| LMCA | — | 29.1% |
| Epoch Capabilities Index | — | 143.98 |
Math Qwen3.5-Flash leads
Qwen3-4B: 29.7 (#240), Qwen3.5-Flash: 37.4 (#158)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 52.2% | 84.4% |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| MathArena Final-Answer Competitions | 38.5% | — |
| LMArena Math | — | 1407 |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-Flash leads
Qwen3-4B: 33.0 (#208), Qwen3.5-Flash: 43.2 (#93)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 52.3% | 82.3% |
| Vectara Hallucination Rate | 5.7% | 10.5% |
| SimpleQA Verified | — | 20.3% |
| LMArena Expert | — | 1407 |
Multilingual Not comparable
Qwen3-4B: —, Qwen3.5-Flash: 50.5 (#121)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | — | 1385 |
| LMArena Chinese | — | 1446 |
| LMArena French | — | 1412 |
| LMArena German | — | 1390 |
| LMArena Japanese | — | 1368 |
| LMArena Korean | — | 1344 |
| LMArena Russian | — | 1379 |
| LMArena Spanish | — | 1400 |
Instruction Following Not comparable
Qwen3-4B: —, Qwen3.5-Flash: 72.6 (#139)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | — | 1374 |
Long Context Not comparable
Qwen3-4B: —, Qwen3.5-Flash: 42.4 (#124)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | — | 1392 |
Writing & Preference Not comparable
Qwen3-4B: —, Qwen3.5-Flash: 57.9 (#122)
| Benchmark | Qwen3-4B | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | — | 1397 |
| LMArena Creative Writing | — | 1343 |
| LMArena Multi-Turn | — | 1393 |
Frequently asked questions
Is Qwen3-4B better than Qwen3.5-Flash?
Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 31.9 on the Noometry Index.
How many benchmarks do Qwen3-4B and Qwen3.5-Flash share?
4 benchmarks have published results for both models. Qwen3-4B has 6 scored results on Noometry and Qwen3.5-Flash has 32.