Program Bench
Moonshot-reported program-generation benchmark used in Kimi K2.7-Code launch comparison tables. Keep separate from public SWE-bench variants.
Models ranked
6
tracked on this benchmark
Score band
77.8 – 26.0
best → lowest tracked
Snapshot trend
-23.30
Jul 16 → Oct 12 · 2 models
Leaderboard
Tracked models ranked by Score (higher is better).
Notes: Model variant: Kimi K3 (max reasoning effort). Benchmark variant: Program Bench. Harness/evaluator: KimiCode. Source posture: Moonshot/Kimi self-reported launch table; confidence: medium because the value is vendor-reported and not independently reproduced. Recommended seed value: 77.8.
Configuration: Muse Wonder Full (max)
Notes: Meta first-party benchmark table on the Muse Wonder model page; methodology in the Muse Wonder model card.
Notes: Moonshot proprietary benchmark; GPT-5.5 score in Moonshot's comparison table should not be mixed with public SWE-bench rows.
Configuration: Muse Wonder (max)
Notes: Meta first-party benchmark table on the Muse Wonder model page; methodology in the Muse Wonder model card.
How to read this benchmark
This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.
Trust this score when
- There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
- The model list covers the same version family you can actually deploy today.
- Top candidates overlap with your required routing and feature requirements.
Be cautious when
- There is only one benchmark snapshot or the dataset appears stale.
- The benchmark metric direction is opposite of your decision objective.
- The score difference between options is narrow and likely within implementation variance.
Related benchmarks
Last reviewed: Jun 19, 2026