Model comparison
Kimi K2.6 vs Qwen3-30B-A3B
Kimi K2.6 is the stronger model overall, scoring 47.7 to 38.9 on the Noometry Index. Qwen3-30B-A3B costs 8.0× less per token, which makes it the better buy when Kimi K2.6'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 K2.6 scores higher in 8 categories and Qwen3-30B-A3B in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.6 leads 57.0 to 37.4.
- The biggest single-benchmark swing is WeirdML: 55.9% for Kimi K2.6 and 29.8% for Qwen3-30B-A3B.
- Qwen3-30B-A3B is cheaper at $0.12 / $0.50 per million input/output tokens, against $0.95 / $4 for Kimi K2.6.
- Kimi K2.6 accepts more context: 262K tokens versus 41K.
Side by side
| Kimi K2.6 | Qwen3-30B-A3B | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 47.7 | 38.9 |
| Released | 2026-04-20 | 2025-04-28 |
| Weights | Open | Open |
| Context window | 262K | 41K |
| Max output | 262K | 16K |
| Input $ / M tokens | $0.95 | $0.12 |
| Output $ / M tokens | $4 | $0.50 |
| Results tracked | 51 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.6 leads
Kimi K2.6: 50.7 (#43), Qwen3-30B-A3B: 37.5 (#194)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| SciCode | 53.5% | 33.3% |
| WeirdML | 55.9% | 29.8% |
| LMArena Coding | 1488 | 1416 |
| SWE-bench Verified | 76.7% | — |
| LMArena WebDev | 1509 | — |
| ALE-Bench | 1,093 | — |
Agentic & Tool Use Qwen3-30B-A3B leads
Kimi K2.6: 21.9 (#137), Qwen3-30B-A3B: 29.8 (#82)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41.4% |
| OSWorld 2.0 | 4.6% | — |
| ExploitBench | 18.4% | — |
| GBAEval | 0.9% | — |
| GDP.pdf | 12% | — |
| Vending-Bench 2 | 6,205 | — |
Reasoning Kimi K2.6 leads
Kimi K2.6: 40.5 (#55), Qwen3-30B-A3B: 22.2 (#204)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| CritPt | 8% | 0.3% |
| Chess Puzzles | 26% | 8% |
| LMArena Hard Prompts | 1470 | 1398 |
| DTBench | 90.9% | 69.3% |
| LMCA | 37.3% | 22.4% |
| Epoch Capabilities Index | 151.05 | 139.63 |
| Kagi LLM Benchmark | — | 54.9% |
| NYT Connections (extended) | 87.2% | — |
| EBR-Bench | 2.4% | — |
| Mystery Game Puzzles | 18% | — |
Math Kimi K2.6 leads
Kimi K2.6: 57.0 (#41), Qwen3-30B-A3B: 37.4 (#157)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| MathArena Final-Answer Competitions | 72.9% | 47.8% |
| OTIS Mock AIME 2024-2025 | 96.1% | 70.3% |
| LMArena Math | 1475 | 1394 |
| FrontierMath (Tiers 1-3) | 57.2% | — |
| FrontierMath Tier 4 | 25.6% | — |
| ProofBench | 16% | — |
| FrontierMath (Feb 2025 set) | 39% | — |
| FrontierMath Tier 4 (v1) | 14.6% | — |
Knowledge Kimi K2.6 leads
Kimi K2.6: 54.0 (#54), Qwen3-30B-A3B: 41.8 (#105)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| GPQA Diamond | 90.8% | 70.1% |
| LMArena Expert | 1491 | 1396 |
| SimpleQA Verified | 34.9% | — |
| Confabulations | — | 12.3% |
| Vectara Hallucination Rate | 10.8% | — |
Multimodal Not comparable
Kimi K2.6: 31.6 (#103), Qwen3-30B-A3B: —
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Vision | 1283 | — |
| Blueprint-Bench 2 | 3.9% | — |
| Furniture Assembly | 21.7% | — |
| LMArena Document | 1451 | — |
Multilingual Kimi K2.6 leads
Kimi K2.6: 54.9 (#37), Qwen3-30B-A3B: 49.5 (#132)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1446 | 1372 |
| LMArena Chinese | 1521 | 1433 |
| LMArena French | 1471 | 1418 |
| LMArena German | 1450 | 1380 |
| LMArena Japanese | 1443 | 1337 |
| LMArena Korean | 1427 | 1331 |
| LMArena Russian | 1446 | 1370 |
| LMArena Spanish | 1464 | 1404 |
Instruction Following Kimi K2.6 leads
Kimi K2.6: 76.3 (#43), Qwen3-30B-A3B: 72.0 (#142)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1451 | 1363 |
Long Context Kimi K2.6 leads
Kimi K2.6: 44.9 (#52), Qwen3-30B-A3B: 31.0 (#283)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1468 | 1379 |
| Fiction.LiveBench | — | 40.6% |
Writing & Preference Kimi K2.6 leads
Kimi K2.6: 68.5 (#26), Qwen3-30B-A3B: 55.6 (#143)
| Benchmark | Kimi K2.6 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1455 | 1384 |
| LMArena Creative Writing | 1434 | 1317 |
| LMArena Multi-Turn | 1453 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
| EQ-Bench Creative Writing | 1725 | — |
| EQ-Bench 4 | 1202 | — |
Frequently asked questions
Is Kimi K2.6 better than Qwen3-30B-A3B?
Kimi K2.6 is the stronger model overall, scoring 47.7 to 38.9 on the Noometry Index. Qwen3-30B-A3B costs 8.0× less per token, which makes it the better buy when Kimi K2.6's lead doesn't matter for your workload.
Which is cheaper, Kimi K2.6 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 K2.6 lists at $0.95 and $4.
Is Kimi K2.6 or Qwen3-30B-A3B better for coding?
Kimi K2.6 scores higher on coding benchmarks: 50.7 versus 37.5 in the Noometry coding category.
Which has the bigger context window?
Kimi K2.6 does, with 262K tokens against 41K.
How many benchmarks do Kimi K2.6 and Qwen3-30B-A3B share?
27 benchmarks have published results for both models. Kimi K2.6 has 51 scored results on Noometry and Qwen3-30B-A3B has 32.