Model detail
Kimi K2.7 Code
Provider: Moonshot · Public slug: moonshotai-kimi-k2-7-code
- Rank
- 13
- Rank movement
- Baseline
- Score
- 39.2
- Last measured
- 2026-07-05
- Coverage
- XHigh not supported; None not supported
Reasoning scores
- XHigh
- —
- Medium
- 39.2
- None
- —
Generation health
Medium0%75%25%n 8 XHigh not supportedNone not supportedHigh repair rateHigh failure rateCost summary
- Mean standard list-price equivalent
- $0.16
- Median unique-cell cost
- $0.14
- Unique-cell range
- $0.11–$0.23
- Coverage
- 8 unique cells / 8 observed runs
- Observed-run total
- $1.28
- Evidence basis
- Provider-reported
- Mean output
- 39k tok
- Output range
- 23k–56k tok
- Token samples
- 8
Cost means equal-weight unique model/game/reasoning cells. Provider-account billed spend is reported separately on Charts.
Run setup
- Overall Models: combines public reasoning variants
- Standard GameBench code-generation prompt