Discover the Best AI Tools

prompt-to-asset creates and packages icons, logos and other visual assets through CLI/MCP, using inline SVG, external prompts or provider APIs. It requires Node.js 20.11+; free or keyless routes depend on provider availability and are not guaranteed unlimited.

media-mcp-server exposes local image resizing, conversion, compression, cropping, filters and metadata removal to AI assistants. Video inspection, frame extraction and conversion require FFmpeg. The tools use FastMCP and Pillow without a service API key; OCR, smart cropping and directory-wide batch processing are listed as roadmap items.

Local MCP server for generating and editing images through user-configured OpenAI-compatible or Gemini image APIs, saving outputs into projects. Requires Node.js 20+, an API endpoint, key and model; it includes no hosted model and upstream charges may apply.

A local Pillow-based MCP server for resizing, cropping, format conversion, compression, watermarks and image overlays. Requires Python and Pillow but no image-service API key. Processing can run offline; these are deterministic editing tools, not a bundled generative image model or a guarantee that the connected assistant is offline.

Provides model discovery, structured prompt building and image, video or audio generation through Pollinations. The README describes keyless free models and paid models requiring a Pollinations key; availability and charges depend on that provider. Local MCP installation does not include the remote models.

An MCP server for multi-provider image generation, editing and video tasks, including interactive region selection and crop-and-stitch repairs. Requires provider credentials and billing. Local installation is not offline inference; automatic Gemini fallback can change the model or resolution after billing errors.

An MCP client for hosted image generation, background removal and upscaling. It can automatically pay x402 charges in USDC on Base using a configured wallet. No API key is required, but the remote operations are paid and this repository does not provide local model inference.

Connects an MCP assistant to browser-based Photopea through WebSocket and JavaScript for image editing and export. Requires a modern browser and a single active editor tab. Refreshing the page can lose unsaved work; the local bridge does not make the hosted editor an offline application.

An MCP client for Seedream image generation, editing and supported layer-decomposition modes through AceDataCloud. Operations are asynchronous and require task polling and platform authentication. Capabilities differ by model: transparent layers and sequential generation are not universal. Open-source client code does not include model weights or free remote usage.

An MCP wrapper for generating, editing and combining images with Nano Banana through AceDataCloud, with single and batch task-status queries. Requires remote platform authentication. Local installation does not make generation offline, and the wrapper’s MIT license does not determine service pricing or model licensing.

An MCP interface to Flux image generation and editing through AceDataCloud, with asynchronous task status and model guidance. Local stdio or hosted HTTP modes require authorized access to the remote platform. The MIT server is a client wrapper, not local Flux inference or a grant of free model-service usage.

Research implementation for inverting real images into StyleGAN using hypernetworks, enabling reconstruction and subsequent editing with tools such as StyleCLIP. Includes training and inference code; documented setup requires pretrained models and an NVIDIA GPU with CUDA, without native CPU support.

Python toolkit for generating images from Russian text with ruDALL-E models, with ruCLIP-based image selection and Real-ESRGAN upscaling. The documented pipeline uses CUDA and downloaded pretrained models; fine-tuning and image-conditioning examples are also provided.

Claude skill for media workflows using AI APIs, URL transformations, content delivery and digital asset management.

ComfyUI is a diffusion-model graphical interface with API and backend using a node-based workflow interface.

Agent-native media processing via Transloadit's 86+ Robots for video encoding, image manipulation, document conversion, OCR, and speech transcription.

Tautulli (Plex media server monitoring) with 11 read-only tools for activity, history, library stats, user stats, transcode analysis, and resolution breakdowns.

Google Maps MCP server with tools for geocode, search, directions, elevation

Mcp-server is an open-source tool for searching and generating images of local businesses worldwide with confidence scores and trust rankings.

MCP server for LightningProx that provides pay-per-request AI access via Bitcoin Lightning, supporting vision/multimodal tasks.

React data visualization MCP server with 30+ chart types

AI-driven demo recording tool that captures browser pages as GIFs or videos with interactions.

An open-source tool that annotates any web page with hover labels for AI assistants using reverse proxy injection without browser extensions.

MCP server for generating screenshots, PDFs, OG images, and narrated video recording from Claude Desktop, Cursor, and Windsurf.