GitHub Copilot Introduces Built-in Natural Language Image Generation and Editing in Public Preview

GitHub has announced a public preview update that allows users to generate and edit images directly within GitHub Copilot using natural language prompts. According to the announcement, the feature utilizes a dedicated image-generation tool powered by GPT Image 2.5 Flare when invoked by a selected reasoning model.

Previously, developers needed to switch contexts to external tools to create visual assets and manually import them into their projects. The new integration aims to streamline this workflow by previewing generated images inline within the chat interface and saving them straight to the workspace.

Key Features and Workflow

• Conversational Creation: Describe desired images using natural language in any Copilot chat session.
• Inline Integration: View generated image previews directly in the chat and save them automatically to the project workspace.
• Editing Capabilities: Modify existing visuals conversationally alongside regular coding tasks.

Availability and Access Requirements

• Supported Interfaces: Accessible via VS Code, the GitHub Copilot app, and the Copilot CLI.
• Access Tier: Currently available in public preview for Copilot Pro, Copilot Pro+, Copilot Business, and Copilot Enterprise subscribers.
• Billing Details: Billed separately from standard chat models, utilizing image tokens that count toward a plan's AI credits.

Administrative Controls and Customization

• Enterprise Policies: Organization and enterprise owners retain the ability to disable image generation by turning off the underlying model through Copilot model policies.
• Watermarking Options: Users can optionally opt in to apply a visible "Made with AI" watermark to generated and edited images, which is disabled by default.

Practical Implications and Verification

This update reduces context switching for developers needing visual assets during software projects. Users should review their organization's specific plan details, billing terms, and administrative policies regarding image tokens and model usage before deployment. To verify the announcement and check for future updates, developers can consult the official GitHub Blog changelog and join discussions in the GitHub Community.