v1.4.7

Just type a character name.Anima images, auto-generated.

One Python script. Runs on Windows right away. Works from iPhone and iPad on the same Wi-Fi. No LLM required.

Enter character
→
LLM generates tags
→
Sent to ComfyUI
→
Image done
Download on GitHub User Guide
Anima Pipeline browser UI and generated images
Browser UI (left) and generated Anima images (right)
Flow

4 steps to generate

Automates the handoff between your browser UI and ComfyUI. Minimal setup to get started.

1
Enter character & scene
Type character details, mood, and style into the UI. Japanese input supported.
Browser UI
2
LLM generates tags
Danbooru-style tags are auto-generated from the character name. Manual editing also available.
Optional
3
Auto-sent to ComfyUI
The prompt is posted to the ComfyUI API, triggering your workflow automatically.
Automated
4
Image ready in browser
Results appear in the in-browser gallery. Save or regenerate with one click.
Gallery
Features

Key features

Built for Windows, with mobile access and extensibility in mind.

AI
LLM tag auto-generation
Supports Gemini API, LM Studio, and any OpenAI-compatible API. Tags generated naturally from character names.
L4
LoRA up to 4 slots
Card UI with thumbnail previews for LoRA selection. Synced with your workflow JSON automatically.
CH
Multi-character management
Save and switch between character presets. Up to 6 characters configurable at once.
NET
Mobile & tablet support
Operate from iPhone / iPad on the same Wi-Fi. ComfyUI requires --listen flag.
UI
JA / EN UI toggle v1.4.7
Auto-detects OS language on first launch. Toggle between Japanese and English anytime.
LOG
Log masking v1.4.7
API keys and sensitive data are masked in logs. Export a ZIP safely for sharing or debugging.
Setup

Up and running in 5 minutes

On Windows, just double-click start_anima_pipeline.bat.

1
Export workflow JSON from ComfyUI
Browse Templates → Anima → Save (API Format), then place the file in workflows/.
2
Run the batch file
Double-click start_anima_pipeline.bat. Dependencies are installed automatically on first run.
3
Open localhost:7860
Open the UI in your browser, set the ComfyUI URL and LLM settings, then save.
4
Enter a character and generate
Press Ctrl+Enter to generate. Results appear in the UI immediately.
localhost:7860
Character name (example input)
Reimu Hakurei, relaxing at shrine LLM
Generated tags (example)
1girl hakurei reimu red bow shrine relaxed
Generate (Ctrl+Enter)
ComfyUI running... 62%
Requirements

What you need

Minimum setup is Python and ComfyUI. LLM is optional.

Python
3.10 or later
Only pip install requests needed. No frameworks, minimal dependencies.
ComfyUI
0.16.4 or later
Requires a working Anima workflow. Add --listen flag for mobile access.
LLM service
Optional
Gemini API / LM Studio / any OpenAI-compatible API supported.
Works without LLM — just enter tags manually

Get started now

Download from GitHub and generate your first image in 5 minutes.
Bug reports via Issue. Latest updates on X.

Download on GitHub User Guide (English) 日本語ページ
note Qiita Zenn X @RHU_AIstudio