Model comparison
GPT-5 Mini vs Qwen3.5 397B-A17B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 41.8 on the Noometry Index. GPT-5 Mini costs 2.0× less per token, which makes it the better buy when Qwen3.5 397B-A17B's lead doesn't matter for your workload.
Last verified . 28 shared benchmarks.
Alibaba (Qwen)Summary
- They share 28 benchmarks with published results for both. GPT-5 Mini scores higher in 2 categories and Qwen3.5 397B-A17B in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5 397B-A17B leads 34.5 to 23.9.
- The biggest single-benchmark swing is Chess Puzzles: 30% for GPT-5 Mini and 13% for Qwen3.5 397B-A17B.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $0.60 / $3.60 for Qwen3.5 397B-A17B.
- GPT-5 Mini accepts more context: 400K tokens versus 262K.
- Qwen3.5 397B-A17B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 41.8 | 46.0 |
| Released | 2025-08-07 | 2026-02-01 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 66K |
| Input $ / M tokens | $0.25 | $0.60 |
| Output $ / M tokens | $2 | $3.60 |
| Results tracked | 60 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 397B-A17B leads
GPT-5 Mini: 40.1 (#146), Qwen3.5 397B-A17B: 42.0 (#114)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Coding | 1406 | 1465 |
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| LMArena WebDev | — | 1400 |
| SWE-bench Multilingual | 39.7% | — |
| SciCode | 39.2% | — |
| WeirdML | 52.7% | — |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Qwen3.5 397B-A17B leads
GPT-5 Mini: 31.1 (#70), Qwen3.5 397B-A17B: 33.3 (#53)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| Terminal-Bench | 34.8% | — |
| APEX-Agents | — | 24.9% |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| τ²-bench Airline | — | 81.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 84.4% |
| τ²-bench Telecom | — | 97.8% |
| Vending-Bench 2 | -31.18 | — |
Reasoning Qwen3.5 397B-A17B leads
GPT-5 Mini: 23.9 (#168), Qwen3.5 397B-A17B: 34.5 (#70)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| Kagi LLM Benchmark | 70.3% | 73.7% |
| Chess Puzzles | 30% | 13% |
| LMArena Hard Prompts | 1380 | 1448 |
| Mystery Game Puzzles | 10% | 18% |
| DTBench | 80.5% | 87.5% |
| LMCA | 34.2% | 37.9% |
| Epoch Capabilities Index | 145.52 | 146.65 |
| ARC-AGI-2 | 4.4% | — |
| NYT Connections (extended) | — | 58.9% |
| ARC-AGI-1 | 54.3% | — |
| CritPt | 0% | — |
| EnigmaEval | 8.2% | — |
| Thematic Generalization | — | 65.1% |
| ForecastBench | 61 | — |
Math Too close to call
GPT-5 Mini: 46.7 (#69), Qwen3.5 397B-A17B: 46.1 (#73)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.7% | 31.2% |
| OTIS Mock AIME 2024-2025 | 86.7% | 88.9% |
| LMArena Math | 1378 | 1454 |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 9% | — |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Qwen3.5 397B-A17B leads
GPT-5 Mini: 45.6 (#86), Qwen3.5 397B-A17B: 53.3 (#58)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 75% | 86.4% |
| LMArena Expert | 1379 | 1462 |
| Humanity's Last Exam | 19.4% | — |
| SimpleQA Verified | 21.6% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Qwen3.5 397B-A17B leads
GPT-5 Mini: 35.6 (#85), Qwen3.5 397B-A17B: 40.7 (#44)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | 1202 | 1263 |
| VPCT | 40.2% | — |
Multilingual Qwen3.5 397B-A17B leads
GPT-5 Mini: 48.9 (#137), Qwen3.5 397B-A17B: 53.7 (#59)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | 1363 | 1430 |
| LMArena Chinese | 1385 | 1500 |
| LMArena French | 1386 | 1461 |
| LMArena German | 1366 | 1447 |
| LMArena Japanese | 1341 | 1426 |
| LMArena Korean | 1308 | 1384 |
| LMArena Russian | 1362 | 1429 |
| LMArena Spanish | 1355 | 1441 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.2 (#46), Qwen3.5 397B-A17B: 75.0 (#77)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | 1357 | 1424 |
| IFEval | 92.7% | — |
Long Context Qwen3.5 397B-A17B leads
GPT-5 Mini: 41.9 (#132), Qwen3.5 397B-A17B: 44.1 (#74)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | 1355 | 1442 |
| Fiction.LiveBench | 69.4% | — |
Writing & Preference Qwen3.5 397B-A17B leads
GPT-5 Mini: 55.2 (#148), Qwen3.5 397B-A17B: 62.3 (#79)
| Benchmark | GPT-5 Mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | 1373 | 1438 |
| LMArena Creative Writing | 1325 | 1401 |
| EQ-Bench Creative Writing | 1313 | 1478 |
| LMArena Multi-Turn | 1363 | 1446 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |
Frequently asked questions
Is GPT-5 Mini better than Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 41.8 on the Noometry Index. GPT-5 Mini costs 2.0× less per token, which makes it the better buy when Qwen3.5 397B-A17B's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Mini or Qwen3.5 397B-A17B?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; Qwen3.5 397B-A17B lists at $0.60 and $3.60.
Is GPT-5 Mini or Qwen3.5 397B-A17B better for coding?
Qwen3.5 397B-A17B scores higher on coding benchmarks: 42.0 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 262K.
How many benchmarks do GPT-5 Mini and Qwen3.5 397B-A17B share?
28 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Qwen3.5 397B-A17B has 36.