Model comparison
Qwen3-4B vs Qwen3.5 397B-A17B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 31.9 on the Noometry Index.
Last verified . 3 shared benchmarks.
Alibaba (Qwen)Summary
- They share 3 benchmarks with published results for both. Qwen3-4B scores higher in 0 categories and Qwen3.5 397B-A17B in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5 397B-A17B leads 53.3 to 33.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 52.2% for Qwen3-4B and 88.9% for Qwen3.5 397B-A17B.
Side by side
| Qwen3-4B | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 46.0 |
| Released | 2025-04-29 | 2026-02-01 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.60 |
| Output $ / M tokens | — | $3.60 |
| Results tracked | 6 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 42.0 (#114)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena WebDev | — | 1400 |
| LMArena Coding | — | 1465 |
Agentic & Tool Use Qwen3.5 397B-A17B leads
Qwen3-4B: 27.6 (#100), Qwen3.5 397B-A17B: 33.3 (#53)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| APEX-Agents | — | 24.9% |
| Berkeley Function Calling Leaderboard | 35.7% | — |
| τ²-bench Airline | — | 81.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 84.4% |
| τ²-bench Telecom | — | 97.8% |
Reasoning Qwen3.5 397B-A17B leads
Qwen3-4B: 19.2 (#268), Qwen3.5 397B-A17B: 34.5 (#70)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| Chess Puzzles | 4% | 13% |
| Kagi LLM Benchmark | — | 73.7% |
| NYT Connections (extended) | — | 58.9% |
| Thematic Generalization | — | 65.1% |
| LMArena Hard Prompts | — | 1448 |
| Mystery Game Puzzles | — | 18% |
| DTBench | — | 87.5% |
| LMCA | — | 37.9% |
| Epoch Capabilities Index | — | 146.65 |
Math Qwen3.5 397B-A17B leads
Qwen3-4B: 29.7 (#240), Qwen3.5 397B-A17B: 46.1 (#73)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 52.2% | 88.9% |
| FrontierMath (Tiers 1-3) | — | 31.2% |
| MathArena Final-Answer Competitions | 38.5% | — |
| LMArena Math | — | 1454 |
Knowledge Qwen3.5 397B-A17B leads
Qwen3-4B: 33.0 (#208), Qwen3.5 397B-A17B: 53.3 (#58)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 52.3% | 86.4% |
| Vectara Hallucination Rate | 5.7% | — |
| LMArena Expert | — | 1462 |
Multimodal Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 40.7 (#44)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | — | 1263 |
Multilingual Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 53.7 (#59)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | — | 1430 |
| LMArena Chinese | — | 1500 |
| LMArena French | — | 1461 |
| LMArena German | — | 1447 |
| LMArena Japanese | — | 1426 |
| LMArena Korean | — | 1384 |
| LMArena Russian | — | 1429 |
| LMArena Spanish | — | 1441 |
Instruction Following Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 75.0 (#77)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | — | 1424 |
Long Context Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 44.1 (#74)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | — | 1442 |
Writing & Preference Not comparable
Qwen3-4B: —, Qwen3.5 397B-A17B: 62.3 (#79)
| Benchmark | Qwen3-4B | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | — | 1438 |
| LMArena Creative Writing | — | 1401 |
| EQ-Bench Creative Writing | — | 1478 |
| LMArena Multi-Turn | — | 1446 |
Frequently asked questions
Is Qwen3-4B better than Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 31.9 on the Noometry Index.
How many benchmarks do Qwen3-4B and Qwen3.5 397B-A17B share?
3 benchmarks have published results for both models. Qwen3-4B has 6 scored results on Noometry and Qwen3.5 397B-A17B has 36.