ComfyUI Lumi Batcher
https://github.com/bytedance/comfyui-lumi-batcher- Category
- AI Tools
- Rank
- No. 994Tools index
- Pricing
- Open Source
- Type
- TOOL
- Builder
- bytedance
- GitHub
- 650 stars
- Date
About
Batch-processing extension for ComfyUI that lets you sweep parameters across runs to speed up workflow debugging instead of adjusting one knob at a time.
What it does
Lumi Batcher adds a visual control surface to ComfyUI for defining grouped parameter combinations, running them across text, image, video, or audio workflows, and browsing the resulting assets in multidimensional grids. Results can be previewed and exported together, while saved images can restore their exact parameters to the canvas.
Why it's ranked here
The strongest case is breadth inside ComfyUI: arbitrary workflow parameters, grouped combinations, multimodal jobs, grid browsing, and asset export live in one interface. Compatibility work for API nodes and established plugins also makes it more practical than a narrow plotting widget. The dependency inconsistencies and tight host coupling temper that appeal.
What's good
Parameter grouping supports meaningful experiments, such as pairing prompts with product-shot settings before crossing them against base models. The result grid keeps outputs visually connected to their combinations. Custom output directories, keyboard cell navigation, single-result downloads, task deletion, and parameter restoration from dragged images reduce the housekeeping around larger runs.
Tradeoffs
This is a ComfyUI extension, not a standalone batch runner. Installation requires Python 3.10 or newer and either a custom-node clone or ComfyUI-Manager. Its Python dependency declarations disagree on several minimum versions, and one source pins pandas while another permits much older releases. The frontend also assumes a specific custom-node location when resolving built assets.
How to use it well
Use it when repeated ComfyUI experiments produce enough variants that manual parameter changes and asset naming become the bottleneck. Define related settings as groups, cross them only where comparisons matter, then inspect and export the grid. It suits multimodal creators and workflow debuggers, but it does not replace ComfyUI or provide a general-purpose standalone automation service.
Technical notes+
__init__.py registers lumi_batcher_service, constructs BatchToolsHandler and CommonHandler, restores package state, attaches task lifecycle hooks, exposes an empty node mapping, and points ComfyUI at ./frontend-setup. prestartup_script.py dynamically loads the service package into sys.modules. frontend-setup/setup.js fetches the built HTML through /api/comfyui-lumi-batcher/get-static-file, imports discovered JavaScript, appends CSS, and registers a ComfyUI extension. frontend/src/index.tsx installs the batch entry point and registers a service worker. frontend/rsbuild.config.ts builds a React frontend with Sass, Less, SVG, and typed CSS-module plugins. pyproject.toml pins pandas==2.2.0, while requirements.txt specifies pandas>=0.25.0 and lower minimums for several other dependencies.
Observed
- License
- GPL-3.0-or-later
- Implementation
- Python service integration with a TypeScript and React frontend
- Install surface
- ComfyUI custom-node clone or installation through ComfyUI-Manager
- Runtime requirement
- Python 3.10 or newer
- Interface
- Integrated ComfyUI panel backed by ComfyUI API routes and a service worker
- Media support
- Text, image, video, and audio batch-processing scenarios
- Platform note
- Repository release notes explicitly describe Windows path compatibility fixes
- Packaging
- Comfy Registry metadata is declared in pyproject.toml
Read from README.md, pyproject.toml, requirements.txt, __init__.py, prestartup_script.py, frontend-setup/setup.js, frontend/eslint.config.mjs, frontend/rsbuild.config.ts, frontend/src/index.tsx, frontend/src/home/index.tsx.
What it can do
Batch process multiple ComfyUI workflows with parameter variations
ComfyUI workflow with parameter ranges → Multiple workflow results with different parameter values
Sweep parameters across multiple runs automatically
Parameter definitions and value ranges → Series of workflow executions with systematically varied parameters
Debug workflows by testing multiple parameter combinations
ComfyUI workflow and parameter test ranges → Comparison results showing workflow behavior across parameter variations
Automate repetitive workflow adjustments
Workflow configuration with multiple parameter sets → Batch of processed results without manual intervention
Speed up workflow optimization through parallel parameter testing
Base workflow and optimization parameter ranges → Performance data across different parameter combinations
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.