Image: GitHub / github.blog Kimi K3 Is Now Available in GitHub Copilot
Moonshot AI's 2.8-trillion-parameter open-weight model is rolling out across Copilot's Pro, Pro+, Max, Business, and Enterprise plans with usage-based billing at competitive rates.
GitHub resumed rolling out Kimi K3 in Copilot on August 6, 2026. The model is an open-weight release from Moonshot AI with 2.8 trillion parameters and a 1-million-token context window. GitHub is hosting it through Fireworks AI and billing it at usage-based rates: $3 per million input tokens, $15 per million output tokens, and $0.30 per million cached input tokens.
The rollout isn’t complete for all users. GitHub is staging it gradually across Copilot Pro, Pro+, Max, Business, and Enterprise. If you’re on Business or Enterprise, an admin needs to flip it on before anyone on the team can select it — Kimi K3 is off by default on those plans.
Where to Find It
Once enabled, the model is selectable from the model picker in VS Code, Visual Studio, the Copilot CLI, GitHub.com, GitHub Mobile (iOS and Android), JetBrains, Xcode, and Eclipse. It sits alongside the other options in Copilot’s growing multi-model lineup.
GitHub’s positioning for Kimi K3 focuses on agentic coding tasks at lower cost. The entry on the changelog describes it as showing “frontier-level abilities on agentic coding with highly cost-effective pricing.” Whether that holds against Claude or GPT in practice depends heavily on the task.
The Open-Weight Caveat
GitHub flags something worth noting for organizations thinking about enabling this: Kimi K3 is an open-weight model, and GitHub’s guidance asks admins to “review open-weight models against their own security, compliance, and data-governance requirements before enabling.” That’s a different governance posture than the proprietary models already in Copilot.
Moonshot AI is the Beijing-based company behind the Kimi chatbot. This isn’t their first model in Copilot — Kimi K2.7 Code went generally available in GitHub Copilot on July 1, 2026. K3 is a larger follow-on with stronger benchmark claims.
Brief Outage Delayed the Launch
The rollout had a rough start. GitHub temporarily paused it on August 6 when a separate GitHub Actions incident caused a 10-hour outage that throttled webhooks and broke Copilot code review. The Kimi K3 rollout was paused as a precaution during that window and resumed afterward.
The addition of Kimi K3 continues a pattern from GitHub: the model picker now includes Claude, GPT, Gemini, Grok, and Kimi options, letting teams swap models per task rather than committing to one provider. For teams already on usage-based Copilot billing, Kimi K3 is available without any plan changes — just the admin toggle if you’re on Business or Enterprise.
Sources: