lidge-ai/ima2-gen

Local-first visual generation runtime and studio for people and coding agents, with reproducible image and video workflows across multiple providers.

What it solves

ima2-gen provides a local-first studio and runtime for generating and managing images and videos across multiple AI providers. It eliminates the need to switch between different web interfaces by consolidating various models into a single local server and interface, while ensuring that all generated assets remain on the user's machine.

How it works

The project runs a small local server that acts as a bridge between the user (via a browser, Mac app, or CLI) and various AI providers. It supports multiple authentication methods, including OAuth for ChatGPT and Grok, and API keys for Gemini, NovelAI, and others. It also integrates with local ComfyUI workflows. All generated files are stored locally in ~/.ima2/generated, and the system tracks metadata like prompts, timing, and settings for every result to ensure reproducibility.

Who it’s for

  • Digital Artists and Designers: Who need a centralized place to generate, branch, and clean up visual assets.
  • Developers and Coding Agents: Who can use the CLI and provided "skills" (Markdown guides) to automate asset pipelines for frontend and UI/UX design.
  • Power Users: Who want to avoid browser-based silos and keep a local history of their AI-generated media.

Highlights

  • Node Graph Branching: Allows users to take a successful image and push it in several different directions without overwriting the original.
  • Canvas Mode: Includes tools for annotation, background removal (via GPT transparency), and exporting to SVG (raster or vector).
  • Multi-Provider Support: Integrates with OpenAI, Grok, Gemini, NovelAI, ComfyUI, and more.
  • Video Production: Supports text-to-video, image-to-video, and a "Storyboard mode" for maintaining character and scene consistency.
  • Agent Integration: Ships with specialized Markdown skills to help coding agents handle asset pipelines and design directions.

Related

  • Project
    kadevin/ilab-conjure

    iLab CONJURE is a locally‑run, FastAPI‑based workstation for generating and editing AI images. It supports multiple back‑ends (GPT‑Image, Google Gemini, Codex) via OpenAI‑compatible APIs or local OAuth, offers a rich web UI with concurrent task queues, history, tagging, prompt‑chip editor, and a command‑line client. Distributed as signed macOS/Windows installers or portable zip, it is AGPL‑3.0 licensed and targets artists, teams, and developers who need a unified, privacy‑first image‑generation workflow.

  • Project
    six-nut/PocketMen-with-you

    A local generative tool that transforms two or more photos into consistent, animated character sprites for Codex companions using open-weight image models.

  • Project
    aqm857886159/Nomi

    Nomi is an open‑source desktop app that lets you generate AI videos by describing a story to an on‑screen agent. The agent plans shots, creates keyframes, renders video using any model you connect (cloud API or local ComfyUI), and assembles everything on a timeline. It runs on macOS and Windows, is licensed AGPL‑3.0, and emphasizes model‑agnostic, pay‑as‑you‑go usage with no built‑in subscription.

  • Project
    CookSleep/gpt_image_playground

    A feature-rich web and desktop interface for managing AI image generation and editing, supporting multiple providers, batch processing, and local-first history management.

  • Project
    hero8152/Infinite-Canvas

    A visual, infinite canvas workspace that integrates multiple AI APIs and local ComfyUI workflows for collaborative generative AI content creation.