Model comparison
Kimi K3 vs Qwen3-30B-A3B
Kimi K3 is the stronger model overall, scoring 59.5 to 38.5 on the Noometry Index. Qwen3-30B-A3B costs 28× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
Alibaba (Qwen)Summary
- They share 27 benchmarks with published results for both. Kimi K3 scores higher in 9 categories and Qwen3-30B-A3B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 21.6.
- The biggest single-benchmark swing is WeirdML: 82.6% for Kimi K3 and 29.8% for Qwen3-30B-A3B.
- Qwen3-30B-A3B is cheaper at $0.12 / $0.50 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 41K.
Side by side
| Kimi K3 | Qwen3-30B-A3B | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 59.5 | 38.5 |
| Released | 2026-07-16 | 2025-04-28 |
| Weights | Open | Open |
| Context window | 1.05M | 41K |
| Max output | 1.05M | 16K |
| Input $ / M tokens | $3 | $0.12 |
| Output $ / M tokens | $15 | $0.50 |
| Results tracked | 53 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Kimi K3: 60.6 (#8), Qwen3-30B-A3B: 37.0 (#188)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| SciCode | 59.5% | 33.3% |
| WeirdML | 82.6% | 29.8% |
| LMArena Coding | 1509 | 1416 |
| DeepSWE | 68.5% | — |
| FrontierCode | 44.2% | — |
| LMArena WebDev | 1659 | — |
| FrontierSWE | 25.9% | — |
| ALE-Bench | 1,524 | — |
Agentic & Tool Use Kimi K3 leads
Kimi K3: 41.8 (#19), Qwen3-30B-A3B: 29.5 (#75)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| APEX-Agents | 50.6% | — |
| Berkeley Function Calling Leaderboard | — | 41.4% |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 32% | — |
| GBAEval | 48.3% | — |
| GDP.pdf | 19% | — |
| Vending-Bench 2 | 5,165 | — |
Reasoning Kimi K3 leads
Kimi K3: 63.0 (#15), Qwen3-30B-A3B: 21.6 (#193)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| CritPt | 23.4% | 0.3% |
| Chess Puzzles | 39% | 8% |
| LMArena Hard Prompts | 1495 | 1398 |
| DTBench | 91.2% | 69.3% |
| LMCA | 52.7% | 22.4% |
| Epoch Capabilities Index | 157.68 | 139.63 |
| ARC-AGI-2 | 60.4% | — |
| SimpleBench | 60.7% | — |
| Kagi LLM Benchmark | — | 54.9% |
| NYT Connections (extended) | 93.6% | — |
| ARC-AGI-1 | 94.5% | — |
| Mystery Game Puzzles | 26% | — |
| Surface Evolver Bench | 95% | — |
| ForecastBench | 61.1 | — |
Math Kimi K3 leads
Kimi K3: 74.0 (#13), Qwen3-30B-A3B: 36.0 (#133)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| MathArena Final-Answer Competitions | 87.8% | 47.8% |
| OTIS Mock AIME 2024-2025 | 97.2% | 70.3% |
| LMArena Math | 1493 | 1394 |
| FrontierMath (Tiers 1-3) | 72.2% | — |
| FrontierMath Tier 4 | 39% | — |
| ProofBench | 87% | — |
Knowledge Kimi K3 leads
Kimi K3: 63.2 (#19), Qwen3-30B-A3B: 41.7 (#96)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| GPQA Diamond | 93.1% | 70.1% |
| LMArena Expert | 1520 | 1395 |
| SimpleQA Verified | 50.6% | — |
| Confabulations | — | 12.3% |
Multimodal Not comparable
Kimi K3: 38.4 (#58), Qwen3-30B-A3B: —
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| Blueprint-Bench 2 | 29.5% | — |
| Furniture Assembly | 34.2% | — |
Multilingual Kimi K3 leads
Kimi K3: 56.5 (#20), Qwen3-30B-A3B: 49.8 (#126)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1466 | 1372 |
| LMArena Chinese | 1528 | 1434 |
| LMArena French | 1488 | 1419 |
| LMArena German | 1474 | 1379 |
| LMArena Japanese | 1492 | 1338 |
| LMArena Korean | 1465 | 1333 |
| LMArena Russian | 1481 | 1369 |
| LMArena Spanish | 1478 | 1403 |
Instruction Following Kimi K3 leads
Kimi K3: 77.8 (#13), Qwen3-30B-A3B: 72.2 (#137)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1481 | 1364 |
Long Context Kimi K3 leads
Kimi K3: 45.8 (#27), Qwen3-30B-A3B: 31.0 (#278)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1494 | 1379 |
| Fiction.LiveBench | — | 40.6% |
Writing & Preference Kimi K3 leads
Kimi K3: 76.7 (#4), Qwen3-30B-A3B: 55.9 (#138)
| Benchmark | Kimi K3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1476 | 1384 |
| LMArena Creative Writing | 1452 | 1318 |
| LMArena Multi-Turn | 1488 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
| EQ-Bench Creative Writing | 2082 | — |
| EQ-Bench 4 | 1339 | — |
Frequently asked questions
Is Kimi K3 better than Qwen3-30B-A3B?
Kimi K3 is the stronger model overall, scoring 59.5 to 38.5 on the Noometry Index. Qwen3-30B-A3B costs 28× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Kimi K3 or Qwen3-30B-A3B?
Qwen3-30B-A3B is cheaper. It lists at $0.12 per million input tokens and $0.50 per million output tokens; Kimi K3 lists at $3 and $15.
Is Kimi K3 or Qwen3-30B-A3B better for coding?
Kimi K3 scores higher on coding benchmarks: 60.6 versus 37.0 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 41K.
How many benchmarks do Kimi K3 and Qwen3-30B-A3B share?
27 benchmarks have published results for both models. Kimi K3 has 53 scored results on Noometry and Qwen3-30B-A3B has 32.