Up to four models
Kimi K2.6 vs Kimi K2.5
On the index (GPT-4 = 100), Kimi K2.6 scores 152 and Kimi K2.5 145. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
Kimi K2.6
152136–169
WeightsOpen
ReleasedApr 2026
Price in / out$0.65 / $3.41
Context262K
Core tests taken4 of 9
Kimi K2.5
145130–161
WeightsOpen
ReleasedJan 2026
Price in / out$0.50 / $2.5
Context262K
Core tests taken5 of 9
Kimi K2.6 and Kimi K2.5 are too close to call (7 points apart, within the uncertainty), ahead on 3 of 3 shared core tests, at 1.3× the price.
Test by testthird-party results
Kimi K2.6Kimi K2.5
Every result as a table
| Test | Kimi K2.6 | Kimi K2.5 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 90.8% | 87.6% |
| GPQA Diamond (Vals AI) | 89.1% | 84.1% |
| MMLU Pro | 87.6% | 85.9% |
| MMMU Pro | 86.3% | 84.3% |
| SimpleQA Verified | 34.9% | 34.3% |
| Maths | ||
| AIME-style maths (OTIS mock) | 96.1% | 92.2% |
| FrontierMath (2025 set) | 39.0% | 27.9% |
| FrontierMath tier 4 (2025 set) | 14.6% | 4.2% |
| Coding | ||
| SWE-bench Verified | 76.6% | 73.8% |
| Code Migration | 27.8% | 7.0% |
| LiveCodeBench | 86.8% | 83.9% |
| SWE-bench | 76.2% | 70.0% |
| Vibe Code Bench v1.1 | 37.9% | 17.5% |
| WeirdML | 55.9% | 45.6% |
| Professional work | ||
| Finance Agent (v2) | 44.9% | 35.8% |
| CorpFin v2 | 66.7% | 68.3% |
| EMB | 57.9% | 28.5% |
| Legal Research Bench | 15.9% | 15.9% |
| MedCode | 40.1% | 39.3% |
| MedScribe | 78.1% | 76.4% |
| MortgageTax | 65.8% | 66.5% |
| TaxEval v2 | 74.7% | 74.2% |
Tests at least two of these models have taken. Bold rows are the core tests.