Up to four models
Qwen3.8-27B vs Kimi K2.6
On the index (GPT-4 = 100), Qwen3.8-27B scores 155 and Kimi K2.6 152. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
155139–171
WeightsOpen
ReleasedAug 2026
Price in / out$0.060 / $2.2
Context1.02M
Core tests taken3 of 9
Kimi K2.6
152136–169
WeightsOpen
ReleasedApr 2026
Price in / out$0.65 / $3.41
Context262K
Core tests taken4 of 9
Qwen3.8-27B and Kimi K2.6 are too close to call (3 points apart, within the uncertainty), ahead on 1 of 1 shared core tests, and costs 2.3× less.
Test by testthird-party results
Every result as a table
| Test | Qwen3.8-27B | Kimi K2.6 |
|---|---|---|
| Reasoning & knowledge | ||
| LiveBench Data Analysis | 76.6% | 65.1% |
| LiveBench IF | 72.7% | 64.4% |
| LiveBench Language | 74.4% | 75.1% |
| LiveBench Reasoning | 80.0% | 79.4% |
| GPQA Diamond (Vals AI) | 88.9% | 89.1% |
| MMLU Pro | 84.3% | 87.6% |
| MMMU Pro | 83.9% | 86.3% |
| Maths | ||
| LiveBench Mathematics | 86.2% | 84.3% |
| Coding | ||
| LiveBench Agentic Coding | 61.4% | 46.9% |
| LiveBench Coding | 75.7% | 78.6% |
| Code Migration | 14.2% | 27.8% |
| LiveCodeBench | 84.0% | 86.8% |
| ProgramBench | 0.0% | 0.0% |
| SWE-bench | 86.0% | 76.2% |
| Vibe Code Bench v1.1 | 64.8% | 37.9% |
| Agents | ||
| DTBench | 88.0% | 90.9% |
| LMCA | 41.4% | 37.3% |
| Professional work | ||
| Finance Agent (v2) | 48.5% | 44.9% |
| EMB | 59.7% | 57.9% |
| LegalBench | 82.4% | 84.7% |
| Legal Research Bench | 36.1% | 15.9% |
| MedCode | 28.7% | 40.1% |
| MedScribe | 83.9% | 78.1% |
| MortgageTax | 64.9% | 65.8% |
| Tax Agent Bench | 30.9% | 12.4% |
| TaxEval v2 | 70.9% | 74.7% |
Tests at least two of these models have taken. Bold rows are the core tests.