Moonshot's Kimi K2.7 Code Is Now GA in GitHub Copilot — First Open-Weight Model in the Picker

TL;DR: Kimi K2.7 Code from Moonshot AI is now GA in GitHub Copilot — 1T-parameter MoE, MIT-licensed, and the first open-weight model in any major IDE coding assistant.

As of July 1, Copilot Pro, Pro+, and Max users can select Kimi K2.7 Code from the model picker. The model uses a Mixture-of-Experts architecture (1T total parameters, 32B active per token); the full weights are public on HuggingFace under the MIT license — GitHub runs a hosted copy on Azure. It completes a five-lab Copilot roster (OpenAI, Anthropic, Google, Microsoft, Moonshot AI). Billed at provider list pricing under usage-based billing; Copilot Business and Enterprise organizations require admin opt-in before users can select it.

Why it matters: The first open-weight model in a major IDE assistant changes the audit and cost calculus. For teams on usage-based billing, routing lighter tasks to Kimi K2.7 costs less than frontier models. For orgs where model provenance matters, MIT-licensed weights with a public HuggingFace release answers the "what model are we actually using?" question in a way closed models can't.

Sources: Kimi K2.7 Code GA in GitHub Copilot — GitHub Changelog · Open-Weight AI Enters GitHub Copilot — TechTimes