GitHub Spark in public preview for Copilot Enterprise subscribers 🚀

[ebndev]

Organizations often struggle to turn ideas into working software. Weeks go into specs or mocks that rarely make it to production, and even when they do, they often miss the mark.

GitHub Spark changes that by enabling teams to focus on the what, not the how: validate ideas early with rapid prototypes, get concepts into production that might otherwise stall in the backlog, and collaborate around live applications instead of fragmented docs or mockups.

Built directly into the GitHub platform that millions of developers already rely on, Spark offers a streamlined path from prototype to production. Repositories, Dependabot, Codespaces, and Copilot all work together out of the box, so teams can move faster without leaving the tools they trust. Developers, product managers, designers, and other team members can centralize collaboration around interactive prototypes, break down silos, and create internal tools that make everyday work more efficient.

Key features

• Natural language to app: Describe your idea and watch Spark build it, with frontend and backend capabilities included.
• No setup required: Data, LLM inference, hosting, deployments, and GitHub auth all included out of the box.
• Add AI to your apps: Add intelligent features powered by LLMs from OpenAI, Meta, DeepSeek, xAI and more—no API key management needed.
• One-click deployments: Publish your app with a single click, with various controls to make sharing secure for you and your team.
• Build your way: Use natural language, visual editing controls, or code with GitHub Copilot code completions at your fingertips to iterate on your ideas.
• Create a repository: Get a repository with your favorite functionality included in just a click. Everything stays synchronized so you aren't trapped in a sandbox.
• Expand with Copilot agents: Open a codespace directly from Spark to iterate with Copilot agent mode or assign an issue to Copilot coding agent.

Now available for Copilot Enterprise customers

Spark is available today in public preview for all Copilot Enterprise customers. Spark messages use premium requests included in GitHub Copilot plans.

How to enable Spark for your organization

By default, Spark is disabled. Enterprise admins can enable Spark as follows:

1. Navigate to your Enterprise Policies tab in GitHub.com.
2. Select Copilot in the left sidebar.
3. Scroll down to the "Features" section where you will see "Spark".
4. Change the Spark policy setting to Enabled, which will enable Spark for all Copilot Enterprise users associated with your enterprise. Optionally, you can change the Spark policy setting to No Policy and enable it at the organization level, which will enable Spark for only certain organizationa within your enterprise.
5. Once enabled, all Copilot subscribers associated with your enterprise or organization can access Spark at github.com/spark.

Learn more

• View more details on the Spark product page.
• Read the Spark documentation.

Leave a comment!

⭐ Join the discussion below to share your feedback and ask questions ⭐

Disclaimer: The UI for features in public preview is subject to change.

[lcherone]

What is the storage limits of the kv, i.e import { useKV } from '@github/spark/hooks'? Couldn't seem to find any info about storage limits.

[jasonrclark]

The current limits are that a key-value pair cannot be larger than 512kb itself, and the overall KV store for a Spark is limited to 2GB.

[Djiit]

Hey team, is there a plan to support private package registry (even only the Github one to start with) to enable organisations to use e.g. their own design system, or any private JS lib? TBH that would be a deal maker for us.

[jasonrclark]

I haven't heard any direct plans around private registry support on our roadmap @Djiit, although we are definitely discussing how to support folks' design systems and templates. I've made a note in our issues about the private registry angle, thanks for bringing it up.

[ANT0071]

😀

[ANT0071]

😀

[reilandeubank]

Are there plans to document the Spark SDK? I am testing out the LLM features in apps, and cannot seem to find any information about the API's for interacting with them.