From the source
GitHub Copilot can now generate and edit images from natural language inside a chat session, with the selected reasoning model calling a dedicated image-generation tool powered by GPT Image 2.5 Flare.
- Previously, getting a visual asset meant leaving Copilot, describing the need again in a separate tool, and manually bringing the result back into the project; GitHub says the update removes that context switch.
- Copilot previews the generated image inline and saves it directly to the workspace, and users can describe an image conversationally at the start of a request or mid-task.
- The feature is in public preview for Copilot Pro, Copilot Pro+, Copilot Business and Copilot Enterprise, and is billed separately from the chat model based on image tokens that count toward a plan's AI credits.
- Organization and enterprise owners can switch image generation off by disabling the image generation model through Copilot model policies.
- An optional visible "Made with AI" watermark can be applied to generated and edited images, and it is off by default.
