Model comparison
Kimi K2.6 vs Qwen2.5 72B Instruct
Kimi K2.6 is the stronger model overall, scoring 47.7 to 31.9 on the Noometry Index.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. Kimi K2.6 scores higher in 8 categories and Qwen2.5 72B Instruct in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.6 leads 57.0 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 96.1% for Kimi K2.6 and 8.1% for Qwen2.5 72B Instruct.
- Kimi K2.6 is cheaper at $0.95 / $4 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Kimi K2.6 accepts more context: 262K tokens versus 131K.
Side by side
| Kimi K2.6 | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 47.7 | 31.9 |
| Released | 2026-04-20 | 2024-09 |
| Weights | Open | Open |
| Context window | 262K | 131K |
| Max output | 262K | 8K |
| Input $ / M tokens | $0.95 | $1.40 |
| Output $ / M tokens | $4 | $5.60 |
| Results tracked | 51 | 43 |
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), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| WeirdML | 55.9% | 16% |
| LMArena Coding | 1488 | 1292 |
| SWE-bench Verified | 76.7% | — |
| LMArena WebDev | 1509 | — |
| SciCode | 53.5% | — |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
| ALE-Bench | 1,093 | — |
Agentic & Tool Use Too close to call
Kimi K2.6: 21.9 (#137), Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| OSWorld 2.0 | 4.6% | — |
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| ExploitBench | 18.4% | — |
| GBAEval | 0.9% | — |
| GDP.pdf | 12% | — |
| METR Time Horizons | — | 35.8% |
| Vending-Bench 2 | 6,205 | — |
Reasoning Kimi K2.6 leads
Kimi K2.6: 40.5 (#55), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1470 | 1271 |
| DTBench | 90.9% | 62.9% |
| LMCA | 37.3% | 13.4% |
| Epoch Capabilities Index | 151.05 | 129 |
| NYT Connections (extended) | 87.2% | — |
| CritPt | 8% | — |
| Chess Puzzles | 26% | — |
| EBR-Bench | 2.4% | — |
| Mystery Game Puzzles | 18% | — |
| BIG-Bench Hard | — | 79.8% |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math Kimi K2.6 leads
Kimi K2.6: 57.0 (#41), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 96.1% | 8.1% |
| LMArena Math | 1475 | 1283 |
| FrontierMath (Tiers 1-3) | 57.2% | — |
| FrontierMath Tier 4 | 25.6% | — |
| MathArena Final-Answer Competitions | 72.9% | — |
| ProofBench | 16% | — |
| Omni-MATH | — | 33% |
| MATH Level 5 | — | 63.2% |
| FrontierMath (Feb 2025 set) | 39% | — |
| FrontierMath Tier 4 (v1) | 14.6% | — |
Knowledge Kimi K2.6 leads
Kimi K2.6: 54.0 (#54), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| GPQA Diamond | 90.8% | 49.1% |
| LMArena Expert | 1491 | 1245 |
| SimpleQA Verified | 34.9% | — |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| Vectara Hallucination Rate | 10.8% | — |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multimodal Not comparable
Kimi K2.6: 31.6 (#103), Qwen2.5 72B Instruct: —
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| 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), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1446 | 1252 |
| LMArena Chinese | 1521 | 1272 |
| LMArena French | 1471 | 1280 |
| LMArena German | 1450 | 1234 |
| LMArena Japanese | 1443 | 1180 |
| LMArena Korean | 1427 | 1188 |
| LMArena Russian | 1446 | 1264 |
| LMArena Spanish | 1464 | 1256 |
Instruction Following Kimi K2.6 leads
Kimi K2.6: 76.3 (#43), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1451 | 1254 |
| IFEval | — | 80.6% |
Long Context Kimi K2.6 leads
Kimi K2.6: 44.9 (#52), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1468 | 1282 |
Writing & Preference Kimi K2.6 leads
Kimi K2.6: 68.5 (#26), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | Kimi K2.6 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1455 | 1269 |
| LMArena Creative Writing | 1434 | 1221 |
| LMArena Multi-Turn | 1453 | 1272 |
| EQ-Bench Creative Writing | 1725 | — |
| WildBench | — | 80.2% |
| EQ-Bench 4 | 1202 | — |
Frequently asked questions
Is Kimi K2.6 better than Qwen2.5 72B Instruct?
Kimi K2.6 is the stronger model overall, scoring 47.7 to 31.9 on the Noometry Index.
Which is cheaper, Kimi K2.6 or Qwen2.5 72B Instruct?
Kimi K2.6 is cheaper. It lists at $0.95 per million input tokens and $4 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Kimi K2.6 or Qwen2.5 72B Instruct better for coding?
Kimi K2.6 scores higher on coding benchmarks: 50.7 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Kimi K2.6 does, with 262K tokens against 131K.
How many benchmarks do Kimi K2.6 and Qwen2.5 72B Instruct share?
23 benchmarks have published results for both models. Kimi K2.6 has 51 scored results on Noometry and Qwen2.5 72B Instruct has 43.