BPM4B

Professional Multimedia Processing Suite

Convert MP3 ↔ M4B · Generate AI Audiobooks · Process Documents to Audio

npm install bpm4b
# Node.js

pip install bpm4b
# Python

What's New

| Version | Feature | | ------- | ------------------------------------------------------------------------------------------------- | | v12 | ⚡ 5x faster processing · EPUB to Audiobook · 50+ doc formats · Google Colab · 99.8%+ TTS coverage OCR Pdfs but will be introduced in v13 | | v11 | ✍️ Interactive Pro Editor · Neural Narration Studio · Multi-voice dialogue | | v10 | 📚 M4B → MP3 · Document to Audiobook · Audio Format Converter |

Installation

npm

npm install bpm4b
cd bpm4b

PyPI

pip install bpm4b
cd bpm4b

Global CLI

npm install -g bpm4b

# Development
npm link

Update

npm update -g bpm4b

Premium Narration (Abogen)

Abogen is the recommended TTS engine powered by Kokoro-82M.

Install espeak-ng

https://github.com/espeak-ng/espeak-ng/releases

Option 1 — Install using uv (Recommended)

# NVIDIA CUDA 12.8 — recommended
uv tool install --python 3.12 abogen[cuda] \
  --extra-index-url https://download.pytorch.org/whl/cu128 \
  --index-strategy unsafe-best-match

# NVIDIA CUDA 12.6 — older drivers
uv tool install --python 3.12 abogen[cuda126] \
  --extra-index-url https://download.pytorch.org/whl/cu126 \
  --index-strategy unsafe-best-match

# NVIDIA CUDA 13.0 — newer drivers
uv tool install --python 3.12 abogen[cuda130] \
  --extra-index-url https://download.pytorch.org/whl/cu130 \
  --index-strategy unsafe-best-match

# CPU only / AMD on Windows
uv tool install --python 3.12 abogen

mkdir abogen
cd abogen

python -m venv venv

# CMD
venv\Scripts\activate.bat

# PowerShell
venv\Scripts\activate.ps1

# NVIDIA GPUs
pip install torch==2.8.0+cu128 torchvision==0.23.0+cu128 torchaudio==2.8.0 \
  --index-url https://download.pytorch.org/whl/cu128

For more install methods, see INSTALL.md

brew install espeak-ng

# Apple Silicon
uv tool install --python 3.13 abogen \
  --with "kokoro @ git+https://github.com/hexgrad/kokoro.git,numpy<2"

# Intel
uv tool install --python 3.12 abogen \
  --with "kokoro @ git+https://github.com/hexgrad/kokoro.git,numpy<2"

For additional setup methods including pip, see INSTALL.md

Usage

Web Interface

bpm4b web

# Custom port
bpm4b web --port 8080

# Custom host
bpm4b web --host 127.0.0.1 --debug

Default URL:

http://localhost:5000

Google Colab

# Auto-detect tunnel
bpm4b web --enable-tunnel

# LocalTunnel
bpm4b web --enable-tunnel --tunnel-service localtunnel

# ngrok
bpm4b web --enable-tunnel --tunnel-service ngrok

See COLAB_USAGE.md for the full Colab setup guide.

CLI — MP3 to M4B

Basic

bpm4b convert input.mp3 output.m4b

Chapters (Seconds)

bpm4b convert book.mp3 book.m4b \
  --chapter "Prologue" 0 \
  --chapter "Chapter 1" 300 \
  --chapter "Chapter 2" 1800

Chapters (MM:SS)

bpm4b convert book.mp3 book.m4b \
  --chapter "Prologue" "0:00" \
  --chapter "Chapter 1" "5:00" \
  --chapter "Chapter 2" "30:00"

Mixed Formats

bpm4b convert book.mp3 book.m4b \
  --chapter "Intro" 0 \
  --chapter "Chapter 1" "6:30" \
  --chapter "Chapter 2" 3600

Help

bpm4b --help
bpm4b web --help
bpm4b convert --help

Chapter timestamps accept:

• Integers (seconds)
• "MM:SS"
• "MM:SS.f"

npm Scripts

npm start

npm run web

npm run convert -- input.mp3 output.m4b

Programmatic Usage (Node.js)

const { convertMp3ToM4b } = require('bpm4b');

await convertMp3ToM4b(
  'input.mp3',
  'output.m4b',
  [
    {
      title: 'Chapter 1',
      start_time: 0
    },
    {
      title: 'Chapter 2',
      start_time: 3600
    }
  ]
);

Features

✍️ Interactive Pro Editor (v11)

• Full text viewer for editing chapter content

• Manual manifest controls

• Merge chapters instantly

• Rename titles

• Exclude segments with one click

• Dual boundary verification

• Universal document support:

  • PDF
  • EPUB
  • DOCX
  • TXT
  • Markdown
  • More

🎙️ Neural Narration Studio (v11)

• Automatic chapter announcements
• Multi-voice dialogue generation
• Narrative vs dialogue voice separation
• Powered by Kokoro-82M

⚡ Performance (v12)

Parallel audio processing with configurable concurrency.

| Method | 100 Files | | -------------------- | ----------- | | Sequential | ~10 min | | Parallel (8 workers) | ~2 min | | Fast Mode + Parallel | ~1.5 min | | TTS Synthesis | 2–3x faster | | Text Coverage | 99.8%+ |

🎙️ EPUB to Audiobook (v12)

Abogen Engine (Default)

curl -X POST http://localhost:5000/api/epub-to-audiobook \
  -F "[email protected]" \
  -F "voice=default" \
  -F "language=en" \
  -F "speed=1.0"

BPM4B Engine

curl -X POST http://localhost:5000/api/epub-to-audiobook \
  -F "[email protected]" \
  -F "voice=af_sky" \
  -F "language=en" \
  -F "speed=1.0" \
  -F "engine=bpm4b"

Engine Comparison

| Feature | Abogen | BPM4B | | ---------------- | -------------- | ------------------- | | Stability | ✅ Reliable | ⚠️ Inconsistent | | Speed | Standard | ⚡ Faster | | Auto-install | ✅ Yes | — | | Parallel workers | — | ✅ Up to 8 | | Voices | 50+ Kokoro Voices | 50+ Kokoro Voices | | Memory Usage | Standard | Lower | | Recommended | ✅ Yes | Speed-critical only |

BPM4B engine may be faster for large files but can produce inconsistent results. Abogen is recommended for reliability.

Supported Languages & Voices

| Language | Voices | | --------------------- | ------ | | 🇺🇸 American English | 20 | | 🇬🇧 British English | 8 | | 🇨🇳 Chinese | 8 | | 🇯🇵 Japanese | 5 | | 🇮🇳 Hindi | 4 | | 🇧🇷 Portuguese | 3 | | 🇪🇸 Spanish | 3 | | 🇮🇹 Italian | 2 | | 🇫🇷 French | 1 |

Playback speed range:

0.5x → 2.0x

📄 Document to EPUB (v12)

curl -X POST http://localhost:5000/api/document-to-epub \
  -F "[email protected]" \
  -F "title=My Book" \
  -F "author=Author Name" \
  -F "language=en"

Documents

• PDF
• DOCX
• DOC
• DOCM
• DOT
• DOTX
• ODT
• ODM
• OTT
• ABW
• WPD

Text

• TXT
• TEXT
• ASC
• ANSI
• LOG
• ME
• MD
• MARKDOWN
• RTF
• CSV

Web

• HTML
• HTM
• XHTML
• XHT
• XML

Academic

• TEX
• BIB

• Chapter 1
• CHAPTER I
• Part 1
• Book 1
• Volume 1
• Markdown headers (##, ###)
• HTML headings (h1, h2, h3)
• Roman numerals (I, II, III)
• Numbered lists (1. Title)

🗂️ File Conversion

| Feature | Description | Since | | ---------------------- | ----------------------------------- | ----- | | MP3 → M4B | Embedded chapters | v1 | | M4B → MP3 | High-fidelity extraction | v10 | | Document → Audiobook | PDF/Text via Kokoro-82M | v10 | | Audio Format Converter | MP3/WAV/FLAC/AAC/OGG/ALAC | v10 | | Folder → M4B | Batch conversion with auto chapters | v10 | | EPUB → Audiobook | Multi-language dual-engine support | v12 |

Audio Format Converter

Supported conversions:

• MP3 ↔ WAV
• FLAC → MP3
• AAC → OGG
• OGG → WAV
• ALAC → FLAC

Supported quality:

• 128k
• 192k
• 256k
• 320k
• Lossless

🎨 Theme System (v11)

25+ themes with real-time switching.

Included Themes

• Dark
• Light
• Matrix
• Cyberpunk
• Dracula
• Monokai
• Vaporwave
• Emerald Forest
• Purple Galaxy
• Sunset Orange
• Blue Ocean
• Cherry Blossom
• Golden Hour
• Midnight Depth
• Royal Velvet
• Arctic Frost
• Volcanic Ash
• Coffee House
• Leafy Greens
• Ocean Breeze
• Lavender Dream
• Steel City
• Ruby Red
• Solarized Light
• High Contrast

📝 Metadata Editor

• Edit title
• Edit author
• Edit genre
• Edit description
• Auto-fill from Open Library
• Upload and embed cover art
• One-click apply + download

🎤 Voice Cloning — KokoClone

• Upload a 3–10 second voice sample

• Synthesize text

• Re-voice existing audio

• Multi-language support:

  • English
  • Hindi
  • French
  • Japanese
  • Chinese
  • Italian
  • Portuguese
  • Spanish

Powered by:

• Kokoro-ONNX
• Kanade voice conversion

🔀 Batch Merge — Audio Glue

• Upload multiple MP3s
• Merge sequentially
• Add metadata
• Add cover art
• Download a single combined file

⏱️ Automatic Chapter Builder

• Enter title + duration
• Automatic timestamp generation
• HH:MM:SS support
• Batch import/export
• Real-time preview
• Minutes or seconds toggle

🌐 Google Colab Defaults (v12)

| Setting | Value | | ------------------ | ------------- | | Concurrency | All CPU cores | | Fast Mode | Enabled | | Audio Quality | 128k | | Max Parallel Files | 16 |

⚙️ UI & Settings

• 25+ real-time themes
• Glassmorphism design
• Drag-and-drop upload
• SSE progress bars
• Live terminal logs
• Copy-to-clipboard commands
• FFmpeg command preview

Web Interface Guide

| Tool | Usage | | --------------- | -------------------------------------- | | MP3 → M4B | Upload MP3 and add chapters | | M4B → MP3 | Convert M4B/M4A to MP3 | | Audiobook Gen | Upload PDF/Text and generate narration | | Metadata Editor | Edit metadata and embed covers | | KokoClone | Voice cloning and synthesis | | Audio Glue | Merge multiple MP3 files |

Advanced — Folder to M4B

const { folderToM4b } = require('bpm4b');

await folderToM4b(
  '/path/to/folder',
  '/path/to/output.m4b',
  {
    concurrency: 8,
    fastMode: true,
    audioQuality: '128k',

    metadata: {
      title: 'My Audiobook',
      author: 'Author Name',
      genre: 'Audiobook'
    },

    onProgress: (percent, msg) => {
      console.log(`${percent}% - ${msg}`);
    }
  }
);

API Reference

POST /api/mp3-to-m4b

| Field | Type | Description | | ---------- | ---- | ---------------------- | | mp3_file | file | Input MP3 | | chapters | JSON | Optional chapter array |

Example:

[
  {
    "title": "Chapter 1",
    "start_time": 0
  },
  {
    "title": "Chapter 2",
    "start_time": "6:30"
  },
  {
    "title": "Chapter 3",
    "start_time": 3600
  }
]

POST /api/convert

| Field | Type | Description | | --------------- | ------ | ----------------- | | source_file | file | MP3 or M4B | | output_name | string | Custom filename | | audio_quality | string | Bitrate | | chapters | JSON | Optional chapters |

POST /api/generate-audiobook

| Field | Type | Description | | ------------- | ------ | --------------- | | doc_file | file | PDF or Text | | voice | string | Kokoro voice | | output_name | string | Output filename |

POST /api/epub-to-audiobook

| Field | Type | Description | | --------------- | ------ | ------------------- | | document_file | file | EPUB file | | voice | string | Voice name | | language | string | Language code | | speed | float | 0.5–2.0 | | engine | string | abogen or bpm4b |

POST /api/document-to-epub

| Field | Type | Description | | --------------- | ------ | ------------------ | | document_file | file | Supported document | | title | string | Book title | | author | string | Author | | language | string | Language code |

POST /api/metadata/extract

| Field | Type | Description | | ------ | ---- | -------------- | | file | file | M4B/M4A source |

Example response:

{
  "title": "Book Title",
  "author": "Author Name",
  "genre": "Fiction",
  "description": "Book description...",
  "coverBase64": "data:image/jpeg;base64,..."
}

POST /api/metadata/apply

| Field | Type | Description | | -------------- | ------ | -------------------- | | file | file | M4B/M4A source | | metadata | JSON | Metadata object | | cover_base64 | string | Optional cover image |

POST /api/convert-audio

| Field | Type | Description | | --------------- | ------ | ------------------------- | | file | file | Audio source | | target_format | string | mp3/wav/flac/aac/ogg/alac | | quality | string | Bitrate or lossless | | job_id | string | Optional SSE job ID |

GET /api/health

Returns:

• Service health status
• FFmpeg availability

Project Structure

bpm4b/
├── bin/
│   └── bpm4b.js
├── lib/
│   ├── core.js
│   ├── server.js
│   ├── audiobook-builder.js
│   └── chapter-detector.js
├── templates/
│   └── index.ejs
├── api/
│   └── index.js
├── uploads/
├── outputs/
├── vercel.json
└── package.json

Deploying to Vercel

1. Push project to GitHub
2. Import into Vercel
3. Deploy

vercel.json is auto-detected.

Important

Serverless functions may time out:

• Hobby: 10s
• Pro: 60s

For large files:

• Use bpm4b web
• Deploy on a dedicated server
• Or upgrade Vercel plan

Notes

• Max upload size: 2GB
• Real-time SSE progress tracking
• Kokoro AI runs locally
• No API keys required
• No external costs
• FFmpeg bundled
• Windows / macOS / Linux support

Typical M4B output size:

0.96GB – 2GB per hour

Depends on bitrate.

Contact

X (Twitter)

@jdjchelp

License

MIT License