Orchestrate complete 3D pipelines — text-to-image, image editing, mesh generation, UV unwrapping, texturing, rigging and animation — then sculpt, assemble, grow trees, raise buildings and author particle VFX in purpose-built editors. All in one workspace, powered by ComfyUI and external APIs.
Start a project as a Kanban board, a node Graph or a Batch run — and browse everything they produce in one shared Assets library.
Manage your 3D assets as cards flowing through automated pipeline stages — Images → Image Edit → Mesh Gen → Mesh Edit → Texturing. Each card carries its own ComfyUI workflow parameters, API tags, and custom attributes.
Visualize the relationships between your assets as a node graph. Connect image sources to edit nodes, mesh generators, and export pipelines. Inspect parameters and trigger runs directly from the graph canvas.
Describe a chain of ComfyUI workflows once, then hand it a list. Declare variables, fill them with groups, and every stage runs once per group — three groups across five stages is fifteen executions, queued and tracked for you.
A centralized library for all your images, meshes, and ComfyUI workflows. Browse, filter, and import assets directly into any project. Track versions and link assets across multiple pipelines.
Kanban, Graph or Batch — whatever the project, a Boards button sits in its header. The first click creates one for you.
An infinite canvas attached to your project, built on Excalidraw. Draw shapes and arrows, pin reference images, block out a character sheet or map out a scene — then generate straight onto the canvas and keep the whole thought next to the work it produced.
Pick a ComfyUI workflow or an external API from the side panel, describe what you want, and the result lands on the canvas where you were thinking. Long runs can be cancelled mid-flight.
Every image you generate, drop or paste is saved into the project's Assets library — so a sketch that turns out well is already one click from a mesh.
Boards autosave as you draw, and flush the moment you switch board, leave the page or close the tab. Local-first, like the rest of your project.
Generation is only the first step. Every editor below is built in, works on the assets already in your project, and saves its result as a new version.
Thirteen modes on the same mesh. Sculpt and paint it by hand, project ComfyUI renders onto it for higher texture quality, unwrap and retopologize it automatically, rig it, bake its maps — then check it against real engine budgets before you ship.
A full raster editor for the images feeding your meshes. Crop, resize, levels, blur and sharpen; paint freehand with brushes; remove shadows on the GPU; or mask a region and hand it to a ComfyUI workflow. Seamless mode makes any texture tile without a visible join.
Auto Rig builds the skeleton and skin weights, then retarget a whole clip library onto it through bone mapping. Pull motion from a video with MoCap, generate it from a text prompt with Kimodo, or import your own — preview live, tick the keepers, and embed them as a new mesh version.
Dress a character. Generate the body and its armour separately, from different prompts and different projects, then bring them into one scene at true relative scale. Fit a piece to a region, nudge it with Elastic Grab, and save every fitted piece plus one merged character ready for Auto Rig.
Grow a tree from a seed. Pick a species, re-roll until the shape is right, and take away a game-ready mesh with bark UVs, foliage cards and wind data already baked in. The whole tree is about two kilobytes of numbers — reproducible, diffable, and able to regenerate its own LODs and impostors.
Draw a plan and a node chain builds every storey — real bays, windows that land on whole numbers, and a roof that fits an L-shaped plan without being told about the L. Eleven style packs rebuild the massing, wings and towers merge in with their own roofs, and you can bind your own window models. Exports with a five-level LOD chain regenerated from the graph.
Node-based particle effects — fire, smoke, sparks, magic, weather, impacts. Start from one of 54 presets across 11 categories, layer several systems on a timeline, generate particle sprites with ComfyUI, bake them to a flipbook, and export the finished effect to Unity or Unreal.
One brush library, shared by every editor. Import Photoshop .abr sets and each sample becomes its own variant — then reach for the same brushes when painting a texture in the Image Editor, or painting, sculpting and displacing directly on a mesh.
From the first prompt to a rigged, animated, engine-ready asset — without leaving the app.
Run any ComfyUI workflow directly from a card or node. Pass dynamic parameters, capture outputs, and chain results to the next pipeline stage automatically.
Integrate any REST or GraphQL API. Tag cards with API endpoints and trigger generation via any third-party 3D service.
Visualize each asset as it flows through your pipeline — Images, Image Edit, Mesh Gen, Mesh Edit, Rigging and Texturing, with card states and progress tracking built in.
Switch to graph mode to see asset dependencies, data flows, and workflow connections as a visual node network.
Centralized library for images, meshes, and workflows. Import local files or pull from your last generation run.
Projects are stored locally on disk — Git-syncable, portable across machines, no cloud lock-in ever.
Chain ComfyUI workflows into stages, declare your variables once, and run the whole chain again for every group of values you supply.
Auto Rig generates a skeleton and skin weights, then retarget clip libraries, video MoCap, or text-prompted motion straight onto your mesh.
GLB, OBJ and FBX for Unity and Unreal — with LOD chains, convex collision hulls, and a Game-Ready check on triangles, UVs, scale and pivot.
Drive the studio from any AI agent. Create projects, run workflows, generate and edit meshes, and tag assets — all over the Model Context Protocol.
An installer that sets up ComfyUI and the Python services for you — or host the server on another machine and share one studio across your team.
Each stage feeds the next. Use ComfyUI workflows or external APIs at any step — the results automatically flow to the next card in your board, and a Batch project repeats the whole chain for every set of values you give it.
Start with a text-to-image generation via ComfyUI, import an existing asset, or use an external API. The result lands as a card in the Images column.
Apply image edits — inpainting, normal map generation, background removal — using any ComfyUI workflow. Cards automatically move to the Image Edit column.
Trigger a mesh generation workflow — Trellis2, Hunyuan, Pixal3D locally, or Tripo AI, Tencent Cloud and Hitem3D through their APIs. The resulting asset moves the card to the Mesh Gen column.
Open the Mesh Editor to unwrap and retopologize automatically, then project ComfyUI renders onto the surface, paint it by hand, and bake normal or AO maps.
Run Auto Rig to generate a skeleton and skin weights, then retarget a clip library, a motion captured from video, or a clip generated from a text prompt — and embed the ones you keep into the mesh.
Export as GLB, OBJ or FBX for Unity and Unreal, with LOD chains and collision hulls. Queue multiple exports on the Action Log and track their status in real time.
Install locally, connect to your ComfyUI instance, and ship your first 3D asset in minutes.
git clone https://github.com/visualbruno/3DGenStudio.git
Whether you want to report a bug, suggest a feature, or just say hello — drop a message and we'll get back to you.