Model comparison
Kimi K2 (Jul 2025) vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 40.9 on the Noometry Index.
Last verified . 36 shared benchmarks.
Alibaba (Qwen)Summary
- They share 36 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 2 categories and Qwen3 235B-A22B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 235B-A22B leads 49.5 to 36.9.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 17.9% for Kimi K2 (Jul 2025) and 9.3% for Qwen3 235B-A22B.
- Kimi K2 (Jul 2025) is cheaper at $0.57 / $2.30 per million input/output tokens, against $0.70 / $2.80 for Qwen3 235B-A22B.
- Kimi K2 (Jul 2025) accepts more context: 262K tokens versus 131K.
Side by side
| Kimi K2 (Jul 2025) | Qwen3 235B-A22B | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 40.9 | 43.2 |
| Released | 2025-07-12 | 2025-04 |
| Weights | Open | Open |
| Context window | 262K | 131K |
| Max output | 262K | 16K |
| Input $ / M tokens | $0.57 | $0.70 |
| Output $ / M tokens | $2.30 | $2.80 |
| Results tracked | 42 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Kimi K2 (Jul 2025): 43.1 (#83), Qwen3 235B-A22B: 43.8 (#72)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| Aider Polyglot | 59.1% | 59.6% |
| WeirdML | 42.8% | 41% |
| LMArena Coding | 1399 | 1445 |
| SWE-bench Verified (bash only) | 63.4% | — |
| SciCode | — | 42.4% |
| GSO | 4.9% | — |
| ALE-Bench | 597.5 | — |
Agentic & Tool Use Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 32.2 (#59), Qwen3 235B-A22B: 33.7 (#48)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 59.1% | 52.1% |
| Terminal-Bench | 35.7% | — |
| METR Time Horizons | 59.2% | — |
| Vending-Bench 2 | — | -11.34 |
Reasoning Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 22.7 (#173), Qwen3 235B-A22B: 15.8 (#302)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| SimpleBench | 26.3% | 31% |
| Kagi LLM Benchmark | 64.4% | 69.4% |
| LMArena Hard Prompts | 1384 | 1433 |
| Epoch Capabilities Index | 146.02 | 143.87 |
| ForecastBench | 60.2 | 59.7 |
| ARC-AGI-2 | — | 1.3% |
| ARC-AGI-1 | — | 11% |
| CritPt | — | 0% |
| Chess Puzzles | — | 12% |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LMCA | — | 29.3% |
Math Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 40.1 (#81), Qwen3 235B-A22B: 48.5 (#51)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| Omni-MATH | 65.4% | 71.8% |
| LMArena Math | 1396 | 1431 |
| FrontierMath (Feb 2025 set) | 21.4% | 8.5% |
| FrontierMath Tier 4 (v1) | 0% | 0% |
| OTIS Mock AIME 2024-2025 | — | 86.7% |
| MATH Level 5 | — | 68.9% |
Knowledge Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 36.9 (#153), Qwen3 235B-A22B: 49.5 (#67)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| MMLU-Pro | 81.9% | 84.4% |
| Confabulations | 20.4% | 15.6% |
| Vectara Hallucination Rate | 17.9% | 9.3% |
| GPQA (HELM) | 65.3% | 72.7% |
| LMArena Expert | 1363 | 1460 |
| GPQA Diamond | — | 80.1% |
| SimpleQA Verified | — | 40.4% |
Multilingual Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 49.8 (#125), Qwen3 235B-A22B: 52.5 (#84)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1372 | 1410 |
| LMArena Chinese | 1416 | 1481 |
| LMArena French | 1379 | 1445 |
| LMArena German | 1386 | 1432 |
| LMArena Japanese | 1349 | 1400 |
| LMArena Korean | 1324 | 1391 |
| LMArena Russian | 1384 | 1411 |
| LMArena Spanish | 1383 | 1429 |
Instruction Following Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 71.3 (#151), Qwen3 235B-A22B: 72.8 (#132)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| IFEval | 85% | 83.5% |
| LMArena Instruction Following | 1347 | 1408 |
Long Context Qwen3 235B-A22B leads
Kimi K2 (Jul 2025): 41.2 (#140), Qwen3 235B-A22B: 46.1 (#25)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| Fiction.LiveBench | 66.7% | 75% |
| LMArena Longer Query | 1353 | 1426 |
| CL-bench | 17.6% | — |
Writing & Preference Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 62.5 (#76), Qwen3 235B-A22B: 59.8 (#103)
| Benchmark | Kimi K2 (Jul 2025) | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1380 | 1419 |
| LMArena Creative Writing | 1349 | 1384 |
| Short-Story Creative Writing | 85.6% | 83% |
| EQ-Bench Creative Writing | 1666 | 1366 |
| WildBench | 86.2% | 86.6% |
| LMArena Multi-Turn | 1371 | 1432 |
Frequently asked questions
Is Kimi K2 (Jul 2025) better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 40.9 on the Noometry Index.
Which is cheaper, Kimi K2 (Jul 2025) or Qwen3 235B-A22B?
Kimi K2 (Jul 2025) is cheaper. It lists at $0.57 per million input tokens and $2.30 per million output tokens; Qwen3 235B-A22B lists at $0.70 and $2.80.
Is Kimi K2 (Jul 2025) or Qwen3 235B-A22B better for coding?
They score almost the same on coding (43.1 vs 43.8); test both on your own repository before choosing.
Which has the bigger context window?
Kimi K2 (Jul 2025) does, with 262K tokens against 131K.
How many benchmarks do Kimi K2 (Jul 2025) and Qwen3 235B-A22B share?
36 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and Qwen3 235B-A22B has 49.