Model comparison
GPT-4o mini vs Qwen3.5 397B-A17B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 25.5 on the Noometry Index. GPT-4o mini costs 5.1× 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-4o mini scores higher in 0 categories and Qwen3.5 397B-A17B in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.5 397B-A17B leads 46.1 to 10.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.9% for GPT-4o mini and 88.9% for Qwen3.5 397B-A17B.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.60 / $3.60 for Qwen3.5 397B-A17B.
- Qwen3.5 397B-A17B accepts more context: 262K tokens versus 128K.
- Qwen3.5 397B-A17B has downloadable open weights; the other is API-only.
Side by side
| GPT-4o mini | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 25.5 | 46.0 |
| Released | 2024-07-18 | 2026-02-01 |
| Weights | Proprietary | Open |
| Context window | 128K | 262K |
| Max output | 16K | 66K |
| Input $ / M tokens | $0.15 | $0.60 |
| Output $ / M tokens | $0.60 | $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-4o mini: 22.0 (#335), Qwen3.5 397B-A17B: 42.0 (#114)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Coding | 1290 | 1465 |
| Aider Polyglot | 3.6% | — |
| LMArena WebDev | — | 1400 |
| WeirdML | 11.8% | — |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 43.1% | — |
| BigCodeBench Complete | 57.4% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Qwen3.5 397B-A17B leads
GPT-4o mini: 27.5 (#101), Qwen3.5 397B-A17B: 33.3 (#53)
| Benchmark | GPT-4o 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% |
| BALROG | 17.4% | — |
Reasoning Qwen3.5 397B-A17B leads
GPT-4o mini: 8.7 (#347), Qwen3.5 397B-A17B: 34.5 (#70)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| Kagi LLM Benchmark | 28.8% | 73.7% |
| Chess Puzzles | 0% | 13% |
| LMArena Hard Prompts | 1267 | 1448 |
| Mystery Game Puzzles | 12% | 18% |
| DTBench | 54.4% | 87.5% |
| LMCA | 10.4% | 37.9% |
| Epoch Capabilities Index | 126.56 | 146.65 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 10.7% | — |
| NYT Connections (extended) | — | 58.9% |
| Thematic Generalization | — | 65.1% |
| LiveBench Reasoning | 32.8% | — |
| LiveBench Data Analysis | 50% | — |
| LiveBench | 41.3% | — |
| PIQA | 88.7% | — |
Math Qwen3.5 397B-A17B leads
GPT-4o mini: 10.4 (#314), Qwen3.5 397B-A17B: 46.1 (#73)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 31.2% |
| OTIS Mock AIME 2024-2025 | 6.9% | 88.9% |
| LMArena Math | 1267 | 1454 |
| Omni-MATH | 28% | — |
| LiveBench Math | 36.3% | — |
| MATH Level 5 | 52.6% | — |
| GSM8K | 91.3% | — |
Knowledge Qwen3.5 397B-A17B leads
GPT-4o mini: 17.7 (#284), Qwen3.5 397B-A17B: 53.3 (#58)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 37.7% | 86.4% |
| LMArena Expert | 1235 | 1462 |
| SimpleQA Verified | 8.3% | — |
| MMLU-Pro | 60.3% | — |
| Confabulations | 37.2% | — |
| GPQA (HELM) | 36.8% | — |
| BoolQ | 88.7% | — |
| MMLU | 81.8% | — |
Multimodal Qwen3.5 397B-A17B leads
GPT-4o mini: 25.9 (#122), Qwen3.5 397B-A17B: 40.7 (#44)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | 1066 | 1263 |
| Video-MME | 64.8% | — |
| GeoBench | 64% | — |
| VPCT | 34% | — |
Multilingual Qwen3.5 397B-A17B leads
GPT-4o mini: 42.0 (#199), Qwen3.5 397B-A17B: 53.7 (#59)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | 1266 | 1430 |
| LMArena Chinese | 1265 | 1500 |
| LMArena French | 1297 | 1461 |
| LMArena German | 1272 | 1447 |
| LMArena Japanese | 1216 | 1426 |
| LMArena Korean | 1195 | 1384 |
| LMArena Russian | 1275 | 1429 |
| LMArena Spanish | 1276 | 1441 |
Instruction Following Qwen3.5 397B-A17B leads
GPT-4o mini: 61.9 (#239), Qwen3.5 397B-A17B: 75.0 (#77)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | 1258 | 1424 |
| LiveBench Instruction Following | 56.8% | — |
| IFEval | 78.2% | — |
Long Context Qwen3.5 397B-A17B leads
GPT-4o mini: 39.1 (#186), Qwen3.5 397B-A17B: 44.1 (#74)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | 1289 | 1442 |
Writing & Preference Qwen3.5 397B-A17B leads
GPT-4o mini: 39.5 (#248), Qwen3.5 397B-A17B: 62.3 (#79)
| Benchmark | GPT-4o mini | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | 1286 | 1438 |
| LMArena Creative Writing | 1268 | 1401 |
| EQ-Bench Creative Writing | 873 | 1478 |
| LMArena Multi-Turn | 1285 | 1446 |
| Short-Story Creative Writing | 67.2% | — |
| WildBench | 79.1% | — |
| LiveBench Language | 28.6% | — |
Frequently asked questions
Is GPT-4o mini better than Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 25.5 on the Noometry Index. GPT-4o mini costs 5.1× 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-4o mini or Qwen3.5 397B-A17B?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen3.5 397B-A17B lists at $0.60 and $3.60.
Is GPT-4o mini or Qwen3.5 397B-A17B better for coding?
Qwen3.5 397B-A17B scores higher on coding benchmarks: 42.0 versus 22.0 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 397B-A17B does, with 262K tokens against 128K.
How many benchmarks do GPT-4o mini and Qwen3.5 397B-A17B share?
28 benchmarks have published results for both models. GPT-4o mini has 60 scored results on Noometry and Qwen3.5 397B-A17B has 36.