// AI NEWS · MODEL RELEASE

GitHub Copilot Adds Kimi K2.7 Code, Its First Open-Weight Model in the Picker

On July 1, 2026, GitHub made Moonshot AI's Kimi K2.7 Code generally available in the Copilot model selector — the first open-weight model offered as a picker option, positioned as a lower-cost choice for coding workflows.

2026-07-02 · by Moe Ameen

What happened

GitHub announced on July 1, 2026 that Kimi K2.7 Code, an open-weight coding model from China's Moonshot AI, is generally available in the Copilot model picker. GitHub calls it the first open-weight model offered as a selectable option in Copilot, framing it as a lower-cost choice alongside the closed models from OpenAI, Anthropic, and Google that Copilot already offers. The model is hosted by GitHub on Microsoft Azure and billed at provider list pricing under Copilot's usage-based billing.

The rollout is staged. GitHub said Kimi K2.7 Code is beginning to reach Copilot Pro, Pro+, and Max plans, with expansion to Copilot Business, Enterprise, and additional surfaces over the following weeks. It is selectable in Visual Studio Code 1.127.0 and newer, Visual Studio 17.14.6 and newer, JetBrains 1.9.1-251 and newer, plus Xcode, Eclipse, the Copilot CLI, github.com, and GitHub Mobile. For Business and Enterprise, the model is off by default — a plan administrator has to enable the Kimi K2.7 Code policy in Copilot settings before anyone in the organization can pick it.

Moonshot AI released Kimi K2.7 Code itself earlier, on June 12, 2026, publishing the full weights to Hugging Face under a Modified MIT license. The model is a Mixture-of-Experts design with roughly one trillion total parameters and about 32 billion active per token, a 256K-token context window, and a forced "thinking" (reasoning) mode. Moonshot's own benchmarks show gains over the prior K2.6 on its internal coding suites; as of its release, independent third-party results on public leaderboards were still limited, so the headline numbers should be read as first-party. Direct API pricing from Moonshot is roughly $0.95 per million input tokens (cache-miss) and $4 per million output tokens — the cost basis behind the "lower-cost" framing, though what a Copilot user pays follows GitHub's usage-based rates.

Why it matters for creators

  • Open-weight models are now first-class in mainstream tooling. Copilot putting an open-weight model in the same picker as OpenAI, Anthropic, and Google normalizes them for everyday developers, not just self-hosting hobbyists.
  • Lower per-token cost changes the math on volume. A cheaper coding model makes high-frequency, agentic tasks — the kind that burn tokens — more affordable, which matters for anyone automating repetitive generation work.
  • It is a coding model, not a content tool. Kimi K2.7 Code writes and reasons over code inside an editor; it does not produce social posts, captioned video, or brand-consistent images. The distinction matters for creators evaluating what it actually does.
  • Choice is the theme of 2026 tooling. Model pickers — in Copilot, in Cursor, across the stack — mean the model is becoming a swappable component, and the durable advantage moves to the workflow wrapped around it.
  • Enterprise is gated. Because the model is off by default for Business and Enterprise, adoption inside companies depends on an admin flipping a policy, not on individual developers.

How to act on this with Kompozy

The plain read: Kimi K2.7 Code is a coding model in a code editor. It ships faster, cheaper token generation for developers, and it is a genuine milestone as the first open-weight model in Copilot's picker — but it does not make content. It will not caption a Short, keep a persona's face consistent across a carousel, or schedule anything to a platform. That is the line worth drawing for any creator who sees "new AI model" and wonders whether it changes their content workflow. It doesn't, directly.

Where it connects is the same place every model launch connects: it's a story your audience is talking about, and being early and clear on it is distribution. Drop "GitHub Copilot added Kimi K2.7 Code, its first open-weight model" into Kompozy and fan it into a Text Post for X and LinkedIn, a Quote Graphic pulling the "first open-weight model in the picker" line, a Carousel breaking down what open-weight means for developers, a captioned Persona Short explaining it to camera, and a Blog Article for the long tail — all in your voice, governed by one Persona Brief, published across nine platforms plus your blog and newsletter while the news is fresh. Kompozy is the generation-and-publishing engine that turns a model-picker headline into a week of on-brand content; the model itself just wrote some code.

Quick takeaways

  • July 1, 2026: GitHub made Kimi K2.7 Code generally available in the Copilot model picker — the first open-weight model offered there.
  • Rolling out to Copilot Pro, Pro+, and Max first; Business and Enterprise follow, off by default until an admin enables the policy.
  • Hosted by GitHub on Microsoft Azure, billed at provider list pricing under usage-based billing.
  • Moonshot AI released the model June 12, 2026 — a ~1T-parameter MoE (~32B active), 256K context, forced thinking mode, Modified MIT license.
  • It is a coding model. Kompozy is what turns the news into captioned, on-brand content published across nine platforms.

Frequently asked questions

When did Kimi K2.7 Code become available in GitHub Copilot?

GitHub announced general availability in the Copilot model picker on July 1, 2026. It began rolling out to Copilot Pro, Pro+, and Max plans first, with Copilot Business, Enterprise, and additional surfaces following over the subsequent weeks. Moonshot AI had released the model itself earlier, on June 12, 2026.

What is Kimi K2.7 Code?

It is an open-weight coding model from Moonshot AI — a Mixture-of-Experts design with roughly one trillion total parameters and about 32 billion active per token, a 256K-token context window, and a forced reasoning ("thinking") mode. Moonshot published the weights on Hugging Face under a Modified MIT license, so it can be self-hosted as well as accessed through hosted providers.

How much does Kimi K2.7 Code cost in Copilot?

GitHub hosts the model on Microsoft Azure and bills it at provider list pricing under Copilot's usage-based billing, so the cost follows GitHub's rates rather than a flat fee. For context, Moonshot's direct API lists roughly $0.95 per million input tokens and $4 per million output tokens — cheaper than most frontier closed models, which is the basis for the "lower-cost" framing.

Can Kimi K2.7 Code create social media content?

No. Kimi K2.7 Code is a coding model built to write and reason over code inside an editor. It does not generate captioned video, brand-consistent images, carousels, or scheduled posts. To turn a model launch like this into on-brand content published across platforms, use a content engine such as Kompozy, which generates and publishes across nine social platforms plus blog and email.

Related news

← All AI news · Get started →