Model comparison
Kimi K2 (Jul 2025) vs Qwen3 Max
Qwen3 Max is the stronger model overall, scoring 43.7 to 41.2 on the Noometry Index. Kimi K2 (Jul 2025) costs 2.4× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 2 categories and Qwen3 Max in 6 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 Max leads 48.1 to 37.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 64.4% for Kimi K2 (Jul 2025) and 72.5% for Qwen3 Max.
- Kimi K2 (Jul 2025) is cheaper at $0.57 / $2.30 per million input/output tokens, against $1.20 / $6 for Qwen3 Max.
- Kimi K2 (Jul 2025) has downloadable open weights; the other is API-only.
Side by side
| Kimi K2 (Jul 2025) | Qwen3 Max | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 41.2 | 43.7 |
| Released | 2025-07-12 | 2025-09-23 |
| Weights | Open | Proprietary |
| Context window | 262K | 262K |
| Max output | 262K | 66K |
| Input $ / M tokens | $0.57 | $1.20 |
| Output $ / M tokens | $2.30 | $6 |
| Results tracked | 42 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Kimi K2 (Jul 2025): 42.4 (#102), Qwen3 Max: 43.0 (#93)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1399 | 1456 |
| ALE-Bench | 597.5 | 370.45 |
| SWE-bench Verified (bash only) | 63.4% | — |
| Aider Polyglot | 59.1% | — |
| GSO | 4.9% | — |
| WeirdML | 42.8% | — |
Agentic & Tool Use Not comparable
Kimi K2 (Jul 2025): 32.4 (#64), Qwen3 Max: —
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| Terminal-Bench | 35.7% | — |
| Berkeley Function Calling Leaderboard | 59.1% | — |
| METR Time Horizons | 59.2% | — |
| Vending-Bench 2 | — | 71.56 |
Reasoning Too close to call
Kimi K2 (Jul 2025): 23.3 (#179), Qwen3 Max: 22.6 (#190)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| Kagi LLM Benchmark | 64.4% | 72.5% |
| LMArena Hard Prompts | 1384 | 1448 |
| Epoch Capabilities Index | 146.01 | 142.38 |
| SimpleBench | 26.3% | — |
| NYT Connections (extended) | — | 30.1% |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 5% |
| DTBench | — | 82.1% |
| LMCA | — | 28.3% |
| ForecastBench | 60.2 | — |
Math Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 42.7 (#83), Qwen3 Max: 38.7 (#131)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Math | 1397 | 1446 |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| OTIS Mock AIME 2024-2025 | — | 73.3% |
| Omni-MATH | 65.4% | — |
| MATH Level 5 | — | 97.1% |
| FrontierMath (Feb 2025 set) | 21.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Qwen3 Max leads
Kimi K2 (Jul 2025): 37.3 (#157), Qwen3 Max: 48.1 (#78)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Expert | 1365 | 1455 |
| GPQA Diamond | — | 72.6% |
| SimpleQA Verified | — | 48.7% |
| MMLU-Pro | 81.9% | — |
| Confabulations | 20.4% | — |
| Vectara Hallucination Rate | 17.9% | — |
| GPQA (HELM) | 65.3% | — |
Multilingual Qwen3 Max leads
Kimi K2 (Jul 2025): 49.6 (#130), Qwen3 Max: 53.7 (#62)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1372 | 1429 |
| LMArena Chinese | 1415 | 1478 |
| LMArena French | 1379 | 1449 |
| LMArena German | 1387 | 1463 |
| LMArena Japanese | 1349 | 1397 |
| LMArena Korean | 1325 | 1399 |
| LMArena Russian | 1385 | 1428 |
| LMArena Spanish | 1386 | 1462 |
Instruction Following Qwen3 Max leads
Kimi K2 (Jul 2025): 71.1 (#156), Qwen3 Max: 74.8 (#87)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1348 | 1419 |
| IFEval | 85% | — |
Long Context Too close to call
Kimi K2 (Jul 2025): 41.2 (#145), Qwen3 Max: 41.6 (#134)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| Fiction.LiveBench | 66.7% | 66.7% |
| CL-bench | 17.6% | 14.5% |
| LMArena Longer Query | 1353 | 1438 |
Writing & Preference Too close to call
Kimi K2 (Jul 2025): 62.3 (#78), Qwen3 Max: 62.4 (#76)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 Max |
|---|---|---|
| LMArena Text | 1380 | 1439 |
| LMArena Creative Writing | 1350 | 1402 |
| LMArena Multi-Turn | 1371 | 1446 |
| Short-Story Creative Writing | 85.6% | — |
| EQ-Bench Creative Writing | 1666 | — |
| WildBench | 86.2% | — |
Frequently asked questions
Is Kimi K2 (Jul 2025) better than Qwen3 Max?
Qwen3 Max is the stronger model overall, scoring 43.7 to 41.2 on the Noometry Index. Kimi K2 (Jul 2025) costs 2.4× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Which is cheaper, Kimi K2 (Jul 2025) or Qwen3 Max?
Kimi K2 (Jul 2025) is cheaper. It lists at $0.57 per million input tokens and $2.30 per million output tokens; Qwen3 Max lists at $1.20 and $6.
Is Kimi K2 (Jul 2025) or Qwen3 Max better for coding?
They score almost the same on coding (42.4 vs 43.0); test both on your own repository before choosing.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Kimi K2 (Jul 2025) and Qwen3 Max share?
22 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and Qwen3 Max has 33.