Model comparison
Grok-3 mini vs Qwen3.5 397B-A17B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 41.2 on the Noometry Index.
Last verified . 21 shared benchmarks.
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. Grok-3 mini scores higher in 1 category and Qwen3.5 397B-A17B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5 397B-A17B leads 34.5 to 13.6.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 61.3% for Grok-3 mini and 73.7% for Qwen3.5 397B-A17B.
- Qwen3.5 397B-A17B has downloadable open weights; the other is API-only.
Side by side
| Grok-3 mini | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | xAI | Alibaba (Qwen) |
| Noometry Index | 41.2 | 46.0 |
| Released | 2025-04-09 | 2026-02-01 |
| Weights | Proprietary | Open |
| Context window | — | 262K |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.60 |
| Output $ / M tokens | — | $3.60 |
| Results tracked | 35 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 397B-A17B leads
Grok-3 mini: 40.8 (#131), Qwen3.5 397B-A17B: 42.0 (#114)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Coding | 1379 | 1465 |
| Aider Polyglot | 49.3% | — |
| LMArena WebDev | — | 1400 |
| WeirdML | 42.6% | — |
Agentic & Tool Use Not comparable
Grok-3 mini: —, Qwen3.5 397B-A17B: 33.3 (#53)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| APEX-Agents | — | 24.9% |
| 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
Grok-3 mini: 13.6 (#334), Qwen3.5 397B-A17B: 34.5 (#70)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| Kagi LLM Benchmark | 61.3% | 73.7% |
| LMArena Hard Prompts | 1375 | 1448 |
| Epoch Capabilities Index | 140.35 | 146.65 |
| ARC-AGI-2 | 0.4% | — |
| NYT Connections (extended) | — | 58.9% |
| ARC-AGI-1 | 16.5% | — |
| Chess Puzzles | — | 13% |
| Thematic Generalization | — | 65.1% |
| Mystery Game Puzzles | — | 18% |
| DTBench | — | 87.5% |
| LMCA | — | 37.9% |
Math Qwen3.5 397B-A17B leads
Grok-3 mini: 42.1 (#85), Qwen3.5 397B-A17B: 46.1 (#73)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 88.9% |
| LMArena Math | 1386 | 1454 |
| FrontierMath (Tiers 1-3) | — | 31.2% |
| Omni-MATH | 31.8% | — |
| MATH Level 5 | 90.9% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
Knowledge Qwen3.5 397B-A17B leads
Grok-3 mini: 46.4 (#81), Qwen3.5 397B-A17B: 53.3 (#58)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 76.3% | 86.4% |
| LMArena Expert | 1395 | 1462 |
| MMLU-Pro | 79.9% | — |
| Confabulations | 10.8% | — |
| GPQA (HELM) | 67.5% | — |
Multimodal Not comparable
Grok-3 mini: —, Qwen3.5 397B-A17B: 40.7 (#44)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | — | 1263 |
Multilingual Qwen3.5 397B-A17B leads
Grok-3 mini: 48.1 (#145), Qwen3.5 397B-A17B: 53.7 (#59)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | 1352 | 1430 |
| LMArena Chinese | 1387 | 1500 |
| LMArena French | 1357 | 1461 |
| LMArena German | 1349 | 1447 |
| LMArena Japanese | 1342 | 1426 |
| LMArena Korean | 1335 | 1384 |
| LMArena Russian | 1353 | 1429 |
| LMArena Spanish | 1381 | 1441 |
Instruction Following Grok-3 mini leads
Grok-3 mini: 78.5 (#9), Qwen3.5 397B-A17B: 75.0 (#77)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | 1357 | 1424 |
| IFEval | 95.1% | — |
Long Context Qwen3.5 397B-A17B leads
Grok-3 mini: 41.0 (#147), Qwen3.5 397B-A17B: 44.1 (#74)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | 1372 | 1442 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Qwen3.5 397B-A17B leads
Grok-3 mini: 52.5 (#169), Qwen3.5 397B-A17B: 62.3 (#79)
| Benchmark | Grok-3 mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | 1370 | 1438 |
| LMArena Creative Writing | 1342 | 1401 |
| LMArena Multi-Turn | 1355 | 1446 |
| Short-Story Creative Writing | 73.5% | — |
| EQ-Bench Creative Writing | — | 1478 |
| WildBench | 65.1% | — |
Frequently asked questions
Is Grok-3 mini better than Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 41.2 on the Noometry Index.
Is Grok-3 mini or Qwen3.5 397B-A17B better for coding?
Qwen3.5 397B-A17B scores higher on coding benchmarks: 42.0 versus 40.8 in the Noometry coding category.
How many benchmarks do Grok-3 mini and Qwen3.5 397B-A17B share?
21 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Qwen3.5 397B-A17B has 36.