Model comparison
Kimi K2 (Jul 2025) vs QwQ-32B
Kimi K2 (Jul 2025) is the stronger model overall, scoring 40.9 to 39.4 on the Noometry Index.
Last verified . 24 shared benchmarks.
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 4 categories and QwQ-32B in 4 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Kimi K2 (Jul 2025) leads 62.5 to 50.9.
- The biggest single-benchmark swing is Aider Polyglot: 59.1% for Kimi K2 (Jul 2025) and 20.9% for QwQ-32B.
Side by side
| Kimi K2 (Jul 2025) | QwQ-32B | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 40.9 | 39.4 |
| Released | 2025-07-12 | 2024-11-28 |
| Weights | Open | Open |
| Context window | 262K | — |
| Max output | 262K | — |
| Input $ / M tokens | $0.57 | — |
| Output $ / M tokens | $2.30 | — |
| Results tracked | 42 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 43.1 (#83), QwQ-32B: 34.9 (#220)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| Aider Polyglot | 59.1% | 20.9% |
| LMArena Coding | 1399 | 1333 |
| SWE-bench Verified (bash only) | 63.4% | — |
| GSO | 4.9% | — |
| WeirdML | 42.8% | — |
| BigCodeBench Instruct | — | 44.6% |
| LiveBench Coding | — | 72.2% |
| BigCodeBench Complete | — | 54.4% |
| ALE-Bench | 597.5 | — |
Agentic & Tool Use Not comparable
Kimi K2 (Jul 2025): 32.2 (#59), QwQ-32B: —
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| Terminal-Bench | 35.7% | — |
| Berkeley Function Calling Leaderboard | 59.1% | — |
| METR Time Horizons | 59.2% | — |
Reasoning Too close to call
Kimi K2 (Jul 2025): 22.7 (#173), QwQ-32B: 22.9 (#171)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| LMArena Hard Prompts | 1384 | 1326 |
| Epoch Capabilities Index | 146.02 | 137.6 |
| ForecastBench | 60.2 | 58.3 |
| SimpleBench | 26.3% | — |
| Kagi LLM Benchmark | 64.4% | — |
| Chess Puzzles | — | 5% |
| LiveBench Reasoning | — | 83.5% |
| LiveBench Data Analysis | — | 65% |
| LiveBench | — | 72% |
Math Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 40.1 (#81), QwQ-32B: 36.2 (#123)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| LMArena Math | 1396 | 1359 |
| OTIS Mock AIME 2024-2025 | — | 59.2% |
| Omni-MATH | 65.4% | — |
| LiveBench Math | — | 77.8% |
| FrontierMath (Feb 2025 set) | 21.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Too close to call
Kimi K2 (Jul 2025): 36.9 (#153), QwQ-32B: 37.2 (#150)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| Confabulations | 20.4% | 15.6% |
| LMArena Expert | 1363 | 1323 |
| GPQA Diamond | — | 65.3% |
| MMLU-Pro | 81.9% | — |
| Vectara Hallucination Rate | 17.9% | — |
| GPQA (HELM) | 65.3% | — |
Multilingual Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 49.8 (#125), QwQ-32B: 45.0 (#171)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| LMArena Non-English | 1372 | 1306 |
| LMArena Chinese | 1416 | 1378 |
| LMArena French | 1379 | 1337 |
| LMArena German | 1386 | 1313 |
| LMArena Japanese | 1349 | 1263 |
| LMArena Korean | 1324 | 1278 |
| LMArena Russian | 1384 | 1297 |
| LMArena Spanish | 1383 | 1353 |
Instruction Following QwQ-32B leads
Kimi K2 (Jul 2025): 71.3 (#151), QwQ-32B: 72.8 (#131)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| LMArena Instruction Following | 1347 | 1297 |
| LiveBench Instruction Following | — | 81.8% |
| IFEval | 85% | — |
Long Context QwQ-32B leads
Kimi K2 (Jul 2025): 41.2 (#140), QwQ-32B: 49.0 (#11)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| Fiction.LiveBench | 66.7% | 83.3% |
| LMArena Longer Query | 1353 | 1308 |
| CL-bench | 17.6% | — |
Writing & Preference Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 62.5 (#76), QwQ-32B: 50.9 (#175)
| Benchmark | Kimi K2 (Jul 2025) | QwQ-32B |
|---|---|---|
| LMArena Text | 1380 | 1329 |
| LMArena Creative Writing | 1349 | 1288 |
| Short-Story Creative Writing | 85.6% | 80.2% |
| EQ-Bench Creative Writing | 1666 | 1257 |
| LMArena Multi-Turn | 1371 | 1314 |
| WildBench | 86.2% | — |
| LiveBench Language | — | 51.4% |
Frequently asked questions
Is Kimi K2 (Jul 2025) better than QwQ-32B?
Kimi K2 (Jul 2025) is the stronger model overall, scoring 40.9 to 39.4 on the Noometry Index.
Is Kimi K2 (Jul 2025) or QwQ-32B better for coding?
Kimi K2 (Jul 2025) scores higher on coding benchmarks: 43.1 versus 34.9 in the Noometry coding category.
How many benchmarks do Kimi K2 (Jul 2025) and QwQ-32B share?
24 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and QwQ-32B has 36.