Model comparison
Qwen3-30B-A3B vs Qwen3 Max
Qwen3 Max is the stronger model overall, scoring 43.7 to 38.9 on the Noometry Index. Qwen3-30B-A3B costs 11× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
Alibaba (Qwen)Summary
- They share 25 benchmarks with published results for both. Qwen3-30B-A3B scores higher in 0 categories and Qwen3 Max in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where Qwen3 Max leads 41.6 to 31.0.
- The biggest single-benchmark swing is Fiction.LiveBench: 40.6% for Qwen3-30B-A3B and 66.7% for Qwen3 Max.
- Qwen3-30B-A3B is cheaper at $0.12 / $0.50 per million input/output tokens, against $1.20 / $6 for Qwen3 Max.
- Qwen3 Max accepts more context: 262K tokens versus 41K.
- Qwen3-30B-A3B has downloadable open weights; the other is API-only.
Side by side
| Qwen3-30B-A3B | Qwen3 Max | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 38.9 | 43.7 |
| Released | 2025-04-28 | 2025-09-23 |
| Weights | Open | Proprietary |
| Context window | 41K | 262K |
| Max output | 16K | 66K |
| Input $ / M tokens | $0.12 | $1.20 |
| Output $ / M tokens | $0.50 | $6 |
| Results tracked | 32 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 Max leads
Qwen3-30B-A3B: 37.5 (#194), Qwen3 Max: 43.0 (#93)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1416 | 1456 |
| SciCode | 33.3% | — |
| WeirdML | 29.8% | — |
| ALE-Bench | — | 370.45 |
Agentic & Tool Use Not comparable
Qwen3-30B-A3B: 29.8 (#82), Qwen3 Max: —
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| Berkeley Function Calling Leaderboard | 41.4% | — |
| Vending-Bench 2 | — | 71.56 |
Reasoning Too close to call
Qwen3-30B-A3B: 22.2 (#204), Qwen3 Max: 22.6 (#190)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| Kagi LLM Benchmark | 54.9% | 72.5% |
| Chess Puzzles | 8% | 4% |
| LMArena Hard Prompts | 1398 | 1448 |
| DTBench | 69.3% | 82.1% |
| LMCA | 22.4% | 28.3% |
| Epoch Capabilities Index | 139.63 | 142.38 |
| NYT Connections (extended) | — | 30.1% |
| CritPt | 0.3% | — |
| Mystery Game Puzzles | — | 5% |
Math Qwen3 Max leads
Qwen3-30B-A3B: 37.4 (#157), Qwen3 Max: 38.7 (#131)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 70.3% | 73.3% |
| LMArena Math | 1394 | 1446 |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| MathArena Final-Answer Competitions | 47.8% | — |
| MATH Level 5 | — | 97.1% |
Knowledge Qwen3 Max leads
Qwen3-30B-A3B: 41.8 (#105), Qwen3 Max: 48.1 (#78)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 70.1% | 72.6% |
| LMArena Expert | 1396 | 1455 |
| SimpleQA Verified | — | 48.7% |
| Confabulations | 12.3% | — |
Multilingual Qwen3 Max leads
Qwen3-30B-A3B: 49.5 (#132), Qwen3 Max: 53.7 (#62)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1372 | 1429 |
| LMArena Chinese | 1433 | 1478 |
| LMArena French | 1418 | 1449 |
| LMArena German | 1380 | 1463 |
| LMArena Japanese | 1337 | 1397 |
| LMArena Korean | 1331 | 1399 |
| LMArena Russian | 1370 | 1428 |
| LMArena Spanish | 1404 | 1462 |
Instruction Following Qwen3 Max leads
Qwen3-30B-A3B: 72.0 (#142), Qwen3 Max: 74.8 (#87)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1363 | 1419 |
Long Context Qwen3 Max leads
Qwen3-30B-A3B: 31.0 (#283), Qwen3 Max: 41.6 (#134)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| Fiction.LiveBench | 40.6% | 66.7% |
| LMArena Longer Query | 1379 | 1438 |
| CL-bench | — | 14.5% |
Writing & Preference Qwen3 Max leads
Qwen3-30B-A3B: 55.6 (#143), Qwen3 Max: 62.4 (#76)
| Benchmark | Qwen3-30B-A3B | Qwen3 Max |
|---|---|---|
| LMArena Text | 1384 | 1439 |
| LMArena Creative Writing | 1317 | 1402 |
| LMArena Multi-Turn | 1378 | 1446 |
| Short-Story Creative Writing | 75.3% | — |
Frequently asked questions
Is Qwen3-30B-A3B better than Qwen3 Max?
Qwen3 Max is the stronger model overall, scoring 43.7 to 38.9 on the Noometry Index. Qwen3-30B-A3B costs 11× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Which is cheaper, Qwen3-30B-A3B or Qwen3 Max?
Qwen3-30B-A3B is cheaper. It lists at $0.12 per million input tokens and $0.50 per million output tokens; Qwen3 Max lists at $1.20 and $6.
Is Qwen3-30B-A3B or Qwen3 Max better for coding?
Qwen3 Max scores higher on coding benchmarks: 43.0 versus 37.5 in the Noometry coding category.
Which has the bigger context window?
Qwen3 Max does, with 262K tokens against 41K.
How many benchmarks do Qwen3-30B-A3B and Qwen3 Max share?
25 benchmarks have published results for both models. Qwen3-30B-A3B has 32 scored results on Noometry and Qwen3 Max has 33.