---
id: "moonshotai/kimi-k2.7-code"
also: ["moonshotai/kimi-k2.7-code-20260612", "moonshotai/Kimi-K2.7-Code", "hf/moonshotai/Kimi-K2.7-Code", "requesty/moonshot/kimi-k2.7-code", "vercel/moonshotai/kimi-k2.7-code"]
card_usd_per_m: "0.95 in, 4.0 out, at moonshotai"
cheapest_list_usd_per_m: "0.6 in, 3.0 out, at coreweave/int4 on openrouter"
cheapest_effective_usd_per_m: "0.2104 in, 3.15 out, at parasail on requesty"
context: 262144
max_output: 235929
input: ["text", "image"]
output: ["text"]
released: "2026-06-12"
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["frequency_penalty", "include_reasoning", "logit_bias", "logprobs", "max_tokens", "min_p", "parallel_tool_calls", "presence_penalty", "reasoning", "repetition_penalty", "response_format", "seed", "stop", "structured_outputs", "temperature", "tool_choice", "tools", "top_k", "top_logprobs", "top_p"]
endpoints: 36
gateways: "openrouter (20), hf (6), requesty (6), vercel (4)"
no_training: 29
trains: 0
retains_prompts: 6
coding: 60.8000
agentic: 21
---
# MoonshotAI: Kimi K2.7 Code

Kimi K2.7 Code is a coding-focused agentic model built on Kimi K2.6, tuned for long-horizon software engineering workflows, multi-step tool use, and end-to-end task completion. It keeps the trillion-parameter MoE architecture with roughly 32B active parameters, a 256K-token context window, native INT4 quantization, and multimodal image and video input support while reducing thinking-token usage compared with K2.6.