Kimi K2.6 vs Kimi K2.7 Code
| Spec | Kimi K2.6 | Kimi K2.7 Code |
|---|---|---|
| Respan Index | 36.3 (#77) | 36.1 (#78) |
| Blended price per 1M | $0.742 | $1.36 |
| Context window | 262K | 262K |
| Input / output per 1M | $0.39 / $1.8 | $0.65 / $3.5 |
| Max output | - | - |
| Input | text, vision | text, vision |
| Released | 2026-04-20 | 2026-06-12 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Kimi K2.6 | Kimi K2.7 Code |
|---|---|---|
| Coding | 13.7 Kimi K2.6 | 12.5 Kimi K2.7 Code |
| Agents | 44.8 Kimi K2.6 | 49.4 Kimi K2.7 Code |
| Reasoning | 40.2 Kimi K2.6 | 36.4 Kimi K2.7 Code |
| Knowledge & language | 35.0 Kimi K2.6 | 36.2 Kimi K2.7 Code |
| Human preference | 53.9 Kimi K2.6 | 52.9 Kimi K2.7 Code |
Benchmarks both report
Kimi K2.6 leads on 13 of the 23 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Kimi K2.6 | Kimi K2.7 Code |
|---|---|---|
| LMArena WebDev (LMArena) | 1509 | 1475 |
| Chess Puzzles (Epoch AI) | 20% | 14% |
| FrontierMath Tier 4 v2 (Epoch AI) | 25.6% | 12.2% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 57.2% | 54% |
| GPQA Diamond | 90.8% | 87.9% |
| LiveBench Agentic Coding (LiveBench) | 46.9% | 45.7% |
| LiveBench Coding (LiveBench) | 78.6% | 74% |
| LiveBench Data Analysis (LiveBench) | 65.1% | 62.7% |
| LiveBench Instruction Following (LiveBench) | 64.4% | 56.3% |
| LiveBench Language (LiveBench) | 75.1% | 77.9% |
| LiveBench Mathematics (LiveBench) | 84.3% | 79.6% |
| LiveBench Reasoning (LiveBench) | 79.4% | 82.8% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 96.1% | 95.6% |
| SimpleQA Verified | 34.9% | 36.5% |
| Vending-Bench 2 (Andon Labs) | 6204.57 | 5082.94 |
| WeirdML (Håvard Tveit Ihle) | 55.9% | 54.1% |
| Toolathlon-Verified (HKUST) | 58% | 58% |
| Kimi Claw 24/7 Bench | 42.9% | 46.9% |
| Kimi Code Bench v2 | 50.9% | 62% |
| MCP-Atlas | 69.4% | 76% |
| MCPMark-Verified | 72.8% | 81.1% |
| MLS-Bench-Lite | 26.7% | 35.1% |
| ProgramBench | 48.3% | 53.6% |
FAQ
Is Kimi K2.6 better than Kimi K2.7 Code?
Kimi K2.6 scores 36.3 on the Respan Index (#77), ahead of Kimi K2.7 Code at 36.1 (#78). Kimi K2.6 leads on 13 of the 23 benchmarks both report.
Which is cheaper, Kimi K2.6 or Kimi K2.7 Code?
Kimi K2.6 lists at $0.39 input and $1.8 output per 1M tokens; Kimi K2.7 Code at $0.65 and $3.5. Kimi K2.6 is 1.8x cheaper per token at list price.
Which has the larger context window?
Both accept 262K tokens of context.