MoonshotAI: Kimi K2.7 Code

Kimi K2.7 Code is a Moonshot AI model that accepts text and image input and supports tool calling and reasoning. It offers a 262,144-token context window with up to 235,929 completion tokens, enough to take in large codebases or long documents and return lengthy outputs in a single pass. The facts available do not confirm structured output support or whether the weights are open, so teams that depend on either should verify before committing. Shortlist it if coding is the priority and output cost matters. At $0.60 per million input tokens and $3.50 per million output tokens, it is priced in a moderate range, and it posts a top coding index score of 100.0 alongside a blended score of 74.1 across three benchmarks. Its agentic score of 48.8 is notably weaker, so for multi-step autonomous workflows it warrants side-by-side testing against alternatives, and coverage of only three benchmarks leaves its broader standing partly unproven.