Nyamuru Manga Name Studio
Open-source PWA for drafting manga storyboards (ネーム) with LLM-ready prompts and a structured OMNY data format.
Japanese-speaking manga creators, doujin authors, and small studios who want a structured, offline-first storyboard editor and are comfortable pairing it with their own LLM of choice.
Users who want a one-click AI manga generator, need an English-language UI out of the box, or expect a finished-art illustration tool rather than a storyboard/name editor.
Nyamuru Manga Name Studio is a free, open-source progressive web app for drafting manga 'names' (ネーム) — the rough storyboard stage that sits between a written script and finished pages. It runs entirely in the browser as a single standalone HTML build, saves work to local storage, and can be used fully offline, which makes it suitable for creators who want a lightweight tool without a cloud account. The editor is built around the author's Nyamuru Data Model (NDM v10) and its serialised counterpart, Open Manga Name YAML (OMNY), so every panel, dialogue balloon, camera direction, and pattern-template layout (A through Q) is captured as structured data you can version, diff, and hand off between tools. A paper-style page preview sits alongside a structured form editor, and OMNY files can be imported and exported for round-tripping. The AI angle — and the reason it belongs in an AI directory — is a bundled set of 'LLM向け生成プロンプト' (LLM-oriented generation prompts) that turn a prose manuscript into a fully-typed NDM v10 / OMNY storyboard, plus documented agent workflows for manga generation and QA. The prompts are model-agnostic: you paste them into ChatGPT, Claude, Gemini, or a local model of your choice, feed in your manuscript, and get back structured YAML you can load straight into the editor. Typical workflows include drafting a short manga chapter from a light-novel scene, iterating panel counts and camera angles with an LLM before committing to layouts, and using the app as a stable canonical editor once the AI-generated draft is in. Because the whole stack is MIT-licensed, studios can fork it, embed it in a pipeline, or ship it as part of a larger tool.
A refreshingly principled take on 'AI for manga': instead of another black-box generator, the author built a real data model (NDM/OMNY) and left the LLM slot open. If you already work in Japanese and live in ChatGPT or Claude, the bundled prompts get you from prose to a clean, editable name faster than any all-in-one product I've tried. The Japanese-only UI and BYO-LLM setup will scare off casual users, but that's the trade-off for something this hackable.
— The AI Tool Bible editorial team
Pros
- ✅ Fully open source under MIT, including code, prompts, docs, and original assets
- ✅ Runs as a single standalone HTML file with offline PWA support and local-storage auto-save
- ✅ Ships a rigorous structured format (NDM v10 / OMNY YAML) so storyboards are diffable and pipeline-friendly
- ✅ Includes ready-made LLM prompts that convert prose manuscripts into typed manga-name YAML
- ✅ Model-agnostic — works with any LLM the creator prefers (ChatGPT, Claude, Gemini, local)
- ✅ Pattern-template library (layouts A–Q) speeds up common page compositions
- ✅ No account, no cloud, no telemetry — usable behind studio firewalls
Cons
- ⚠️ Japanese-only UI and documentation; non-Japanese users will need to translate the interface
- ⚠️ No built-in AI — you must copy prompts into an external LLM and paste OMNY back
- ⚠️ Small solo project (single maintainer, ~10 GitHub stars) with limited support and roadmap visibility
- ⚠️ Not a drawing tool — outputs storyboards and structured data, not finished art or pages
- ⚠️ Local-storage-only persistence means clearing browser data can lose work if not exported
Use cases
Explore related
Compare with similar tools
All in Writing →
Claude
FeaturedAnthropic's flagship assistant for long-form writing, analysis, and coding.

GPT-4o
FeaturedOpenAI's multimodal flagship behind ChatGPT.

Gemini Advanced
Google's flagship — strong at math, long context, and Workspace integration.

Llama 3
Meta's open-weights LLM family that put serious frontier-adjacent models in everyone's hands.

Qwen
Alibaba's open-weight foundation model family covering chat, vision, image generation, translation, and safety classification.

Jan
Open-source desktop ChatGPT alternative that runs local LLMs and routes to cloud providers from one app.