Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-08-16 · high
Microsoft's own MAI-Code-1.1-Flash shipped in the same release, two days before local runtimes could even run K3.
Why now: GitHub's changelog for the August 10 weekly release posted August 13, two days before local open-source runtimes matched K3 support at all.
GitHub rolled Kimi K3 into every Copilot tier in its August 10 weekly release, per the GitHub Changelog. Local open-source runtimes only merged support for the model on August 15, and none has shipped a tagged build yet. Copilot is now the fastest way for most developers to actually run it. Moonshot's open-weight model runs 2.8 trillion parameters across 896 experts, and until this release it sat behind the plan's top tiers.
The same release starts rolling out Microsoft's own MAI-Code-1.1-Flash, with native image understanding. The August 10 drop puts Moonshot's model and Microsoft's own model in front of subscribers at the same time.
Bundling a rival's model into every tier for free, right alongside its own new model, is a bet that distribution beats owning the model. Watch whether Moonshot's weights end up handling more of Copilot's daily coding traffic than MAI-Code. If it does, GitHub's bundling bet looks smarter than owning the model outright.
The changelog doesn't say why K3 skipped a usual top-tier-first rollout. It posted August 13, three days after the August 10 release it describes, and two days before local runtimes matched K3 support at all.
Each link below shares sources, entities, or timing with this story.
On June 18 GitHub made its purpose-built small coding model available across the Copilot app, CLI, Chat, Visual Studio, JetBrains, Eclipse, GitHub Mobile, and Xcode, in Free through Max plans. (GitHub) A fast, cheap first-party model pushed this broadly signals Microsoft routi...
MAI-Code-1-Flash, a 5B-parameter coding model, is in GitHub Copilot and VS Code, and Microsoft says it beats Claude Haiku 4.5 across core coding benchmarks, +16 points on SWE-Bench Pro at 51.2% versus 35.2%, using up to 60% fewer tokens. MAI-Thinking-1, a 35B-active MoE with a...
The update adds native image input and faster token streaming alongside gains in instruction following and tool use. It arrives with Polaris becoming the default model on every Copilot seat and Visual Studio 18.9 exposing low/medium/high thinking-effort controls. RuntimeWire M...
An open-weight Chinese frontier model is now a dropdown option in Microsoft's coding product. That happened before anyone finished characterizing what the model does. GitHub's changelog dated August 6 makes Kimi K3 generally available across Copilot Pro, Pro+, Max, Business an...
GitHub announced that every Copilot plan moves to usage-based billing via "AI Credits" on June 1, 2026. The flat-rate model is dead. Every plan gets a monthly credit allotment calculated by actual token consumption (input, output, cached) at listed API rates. Code completions...
A spec is a press release until someone who didn't write it implements it. GitHub made Agent Plugins 1.0 generally available on August 12 across VS Code, Copilot CLI, the Copilot SDK, and the Copilot app on all plans. The spec, published August 6, was co-authored by AWS, Anysp...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.