Feature walkthrough

A local-first workspace for AI-generated assets

Generate, organize and refine images, 3D models, sound and splats — from a single prompt to a fully-tagged, searchable library. Here's the tour, feature by feature.

Local-first · ComfyUI + Ollama Multi-modal · image · video · sound · 3D · splats Cloud optional · Meshy · Tripo · OpenAI
01Library≈ 5s

Your whole asset workspace at a glance

Every generated image, 3D model, prop and texture in one categorized grid — filter, sort, and open any asset.

localhost:8080/library
The Library view — every asset in one categorized grid
02Detail≈ 5s

Every asset, fully documented

Prompt, tags, measurements and model metadata — then continue straight into Retexture, Remesh or Image → 3D.

localhost:8080/asset
The asset detail view with prompt, tags and metadata
03Generate≈ 7s

One prompt, any medium

Image, video, sound, 3D or Gaussian splats — run it locally through ComfyUI or hand it to a cloud provider, then hit Start.

localhost:8080/generate
The Generate view — one prompt, any medium
04Activity≈ 5s

Keep an overview of what's happening

Every job in one place — queued, generating and completed — so you always know what your pipeline is doing at a glance.

localhost:8080/activity
The Activity view — queued, generating and completed jobs
05Search≈ 5s

Find anything, semantically

Type what you mean, not the exact filename — matching tags and results surface instantly across the library.

localhost:8080/search
The Search view — semantic search across the library
06Settings≈ 5s

Route each task to the model you want

Pick the provider per task — local ComfyUI, Ollama or cloud — and manage workflows, config and users.

localhost:8080/settings
The Settings view — route each task to a provider

That's the tour.

Open source and self-hostable — clone it, run docker compose up, and start creating.