Roadmap

img-mcp is becoming a general-purpose image manipulation MCP server. Generation is one tool; the main focus is fast, local, non-AI image operations powered by Node.js.

Shipped

• [x] Text-to-image generation — create images from natural-language prompts.
• [x] Reference-image generation — guide generation with base64 or data-URI images.
• [x] Multiple aspect ratios and sizes — 1:1, 16:9, 9:16, 4K, and more.
• [x] Transparent backgrounds — remove backgrounds via @imgly/background-removal-node.
• [x] Zero-install usage — run with npx -y img-mcp.
• [x] Bring-your-own-key — your Google API key stays in your environment.
• [x] Resize — scale by width, height, percentage, or fit mode (cover, contain, fill, inside, outside).
• [x] Convert formats — JPEG, PNG, WebP, AVIF, GIF.

Planned — Core image manipulation

The priority is robust, dependency-light Node.js operations that work without an API key.

• [ ] Crop — crop by coordinates/dimensions or gravity-based smart crop.
• [ ] Compress / optimize — quality, format-specific tuning, and target file-size optimization.
• [ ] Rotate & flip — arbitrary rotation, horizontal/vertical flip, auto-orient from EXIF.
• [ ] Blur & sharpen — gaussian blur, unsharp mask, motion blur.
• [ ] Adjustments — brightness, contrast, saturation, hue, exposure, gamma.
• [ ] Filters & effects — grayscale, sepia, invert, tint, rounded corners, drop shadow.
• [ ] Metadata read/write — EXIF, IPTC, XMP extraction and stripping.
• [ ] Batch operations — apply the same transform to multiple images in one call.

Planned — Advanced manipulation

• [ ] Watermark / overlay — add text or image watermarks with positioning and opacity.
• [ ] Collage / montage — combine multiple images into a grid or custom layout.
• [ ] Sprite sheet generation — pack frames into a single image.
• [ ] Image comparison — diff two images and return similarity metrics.
• [ ] Thumbnails — generate preset thumbnail sizes with face/center gravity.
• [ ] Animated image support — split/combine GIF/WebP frames.
• [ ] SVG rasterization — render SVGs to PNG/JPEG at any size.

Research

Longer-term or optional capabilities that may rely on AI or heavier dependencies.

• [ ] Additional generation providers — support image models beyond Google Nano Banana.
• [ ] Upscale & super-resolution — increase resolution and recover detail.
• [ ] Inpainting / outpainting — edit or extend image regions.
• [ ] Background replacement — replace backgrounds via segmentation or color keying.
• [ ] Style transfer & presets — apply consistent artistic styles and brand presets.
• [ ] Image analysis feedback loop — let agents critique and iteratively refine images.

Share your ideas

Have a feature request? Open an issue on GitHub and let us know what you'd like to see next.