Model comparison
GPT-5 Nano vs Qwen3-30B-A3B
Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 33.5 on the Noometry Index. GPT-5 Nano costs 1.6× less per token, which makes it the better buy when Qwen3-30B-A3B's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
Alibaba (Qwen)Summary
- They share 26 benchmarks with published results for both. GPT-5 Nano scores higher in 2 categories and Qwen3-30B-A3B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3-30B-A3B leads 55.6 to 39.1.
- The biggest single-benchmark swing is Chess Puzzles: 27% for GPT-5 Nano and 8% for Qwen3-30B-A3B.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $0.12 / $0.50 for Qwen3-30B-A3B.
- GPT-5 Nano accepts more context: 400K tokens versus 41K.
- Qwen3-30B-A3B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Qwen3-30B-A3B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 33.5 | 38.9 |
| Released | 2025-08-07 | 2025-04-28 |
| Weights | Proprietary | Open |
| Context window | 400K | 41K |
| Max output | 128K | 16K |
| Input $ / M tokens | $0.05 | $0.12 |
| Output $ / M tokens | $0.40 | $0.50 |
| Results tracked | 49 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3-30B-A3B leads
GPT-5 Nano: 33.6 (#254), Qwen3-30B-A3B: 37.5 (#194)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| WeirdML | 38.1% | 29.8% |
| LMArena Coding | 1351 | 1416 |
| SWE-bench Verified (bash only) | 34.8% | — |
| SciCode | — | 33.3% |
| ALE-Bench | 718.67 | — |
Agentic & Tool Use Qwen3-30B-A3B leads
GPT-5 Nano: 25.8 (#106), Qwen3-30B-A3B: 29.8 (#82)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 51.5% | 41.4% |
| Terminal-Bench | 21.8% | — |
Reasoning Qwen3-30B-A3B leads
GPT-5 Nano: 16.3 (#306), Qwen3-30B-A3B: 22.2 (#204)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| Kagi LLM Benchmark | 62.2% | 54.9% |
| Chess Puzzles | 27% | 8% |
| LMArena Hard Prompts | 1328 | 1398 |
| DTBench | 62.7% | 69.3% |
| LMCA | 7.9% | 22.4% |
| Epoch Capabilities Index | 139.38 | 139.63 |
| ARC-AGI-2 | 2.6% | — |
| ARC-AGI-1 | 20.7% | — |
| CritPt | — | 0.3% |
| Mystery Game Puzzles | 9% | — |
| ForecastBench | 59.1 | — |
Math Qwen3-30B-A3B leads
GPT-5 Nano: 29.4 (#241), Qwen3-30B-A3B: 37.4 (#157)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 81.1% | 70.3% |
| LMArena Math | 1317 | 1394 |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| MathArena Final-Answer Competitions | — | 47.8% |
| ProofBench | 12% | — |
| Omni-MATH | 54.6% | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Qwen3-30B-A3B leads
GPT-5 Nano: 35.9 (#178), Qwen3-30B-A3B: 41.8 (#105)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| GPQA Diamond | 69.4% | 70.1% |
| LMArena Expert | 1321 | 1396 |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| Confabulations | — | 12.3% |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal Not comparable
GPT-5 Nano: 31.3 (#108), Qwen3-30B-A3B: —
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| LMArena Vision | 1159 | — |
| VPCT | 37.2% | — |
Multilingual Qwen3-30B-A3B leads
GPT-5 Nano: 45.3 (#172), Qwen3-30B-A3B: 49.5 (#132)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1313 | 1372 |
| LMArena Chinese | 1356 | 1433 |
| LMArena German | 1327 | 1380 |
| LMArena Japanese | 1226 | 1337 |
| LMArena Korean | 1269 | 1331 |
| LMArena Russian | 1296 | 1370 |
| LMArena Spanish | 1360 | 1404 |
| LMArena French | — | 1418 |
Instruction Following GPT-5 Nano leads
GPT-5 Nano: 75.0 (#79), Qwen3-30B-A3B: 72.0 (#142)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1306 | 1363 |
| IFEval | 93.2% | — |
Long Context Too close to call
GPT-5 Nano: 31.3 (#281), Qwen3-30B-A3B: 31.0 (#283)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| Fiction.LiveBench | 44.4% | 40.6% |
| LMArena Longer Query | 1312 | 1379 |
Writing & Preference Qwen3-30B-A3B leads
GPT-5 Nano: 39.1 (#249), Qwen3-30B-A3B: 55.6 (#143)
| Benchmark | GPT-5 Nano | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1320 | 1384 |
| LMArena Creative Writing | 1249 | 1317 |
| LMArena Multi-Turn | 1311 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
Frequently asked questions
Is GPT-5 Nano better than Qwen3-30B-A3B?
Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 33.5 on the Noometry Index. GPT-5 Nano costs 1.6× less per token, which makes it the better buy when Qwen3-30B-A3B's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or Qwen3-30B-A3B?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; Qwen3-30B-A3B lists at $0.12 and $0.50.
Is GPT-5 Nano or Qwen3-30B-A3B better for coding?
Qwen3-30B-A3B scores higher on coding benchmarks: 37.5 versus 33.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 41K.
How many benchmarks do GPT-5 Nano and Qwen3-30B-A3B share?
26 benchmarks have published results for both models. GPT-5 Nano has 49 scored results on Noometry and Qwen3-30B-A3B has 32.