n8n-nodes-aivisspeech

English | 日本語

This is an n8n community node. It lets you use AivisSpeech (VOICEVOX-compatible) text-to-speech synthesis in your n8n workflows.

AivisSpeech is a high-quality Japanese text-to-speech engine with VOICEVOX-compatible API.

n8n is a fair-code licensed workflow automation platform.

Installation Operations Credentials Compatibility Resources

Installation

Follow the installation guide in the n8n community nodes documentation.

Community Nodes (Recommended)

1. Go to Settings > Community Nodes
2. Select Install
3. Enter n8n-nodes-aivisspeech
4. Select Install

Manual Installation

cd ~/.n8n/custom
npm init -y
npm install n8n-nodes-aivisspeech

Operations

Speech Synthesis

| Operation | Description | |-----------|-------------| | Get Speakers | List available speakers and their voice styles | | Synthesize (Simple) | Convert text to speech (auto 2-step: AudioQuery + Synthesis) | | Get AudioQuery | Get AudioQuery (pronunciation/accent info) from text | | Synthesize from AudioQuery | Synthesize speech from AudioQuery JSON | | Multi-Text Synthesis | Synthesize multiple texts and combine into one audio file |

Reference

| Operation | Description | |-----------|-------------| | Get Parameter Guide | Get parameter guide (Markdown), format guide for AI prompts, and JSON Schema |

User Dictionary

| Operation | Description | |-----------|-------------| | Get User Dictionary | List all words in user dictionary | | Add Word | Add a word to user dictionary | | Update Word | Update an existing word in user dictionary | | Delete Word | Delete a word from user dictionary |

Credentials

To use this node, you need to configure the AivisSpeech API credentials:

| Field | Description | Default | |-------|-------------|---------| | Base URL | AivisSpeech engine URL | http://localhost:10101 |

Docker users: Use http://host.docker.internal:10101 if AivisSpeech runs on your host machine.

Compatibility

• n8n version: 1.0.0 or later
• AivisSpeech: Any version with VOICEVOX-compatible API
• VOICEVOX: Compatible with VOICEVOX Engine API

Resources

• n8n community nodes documentation
• AivisSpeech GitHub
• VOICEVOX API Reference

Changelog

0.4.11

• Changed: merged getFormatGuide operation into getParameterGuide (now returns parameterGuide, formatGuide, and jsonSchema in one call)
• Changed: removed system prompt intro from format guide (users provide their own system prompt)

0.4.10

• Added: format guide operation (getFormatGuide) returning AI prompt instruction Markdown for JSON generation
• Changed: format definition converted to standard JSON Schema 2020-12
• Changed: all UI parameter labels now show bilingual format (e.g., "話速: speedScale")
• Changed: JSON input hint formatted with monospace styling and gray comments

0.4.9

• Added: parameter guide operation (getParameterGuide) returning detailed parameter guide (Markdown) and JSON Schema
• Added: docs/parameter-guide.md standalone documentation

0.4.8

• Changed: intonationScale label renamed from "抑揚" to "感情表現" to match AivisSpeech API semantics
• Added: min/max values to all parameter descriptions
• Added: JSON input hint with formatted full-set example below input field

0.4.7

• Added: pauseLength and tempoDynamicsScale parameters to all synthesis operations
• Added: full-set JSON sample in multi-text synthesis JSON input default
• Fixed: JSON input mode now correctly parses string input (jsonInput.map is not a function)

0.4.6

• Changed: simple synthesis audio params (speed, pitch, intonation, volume, silence) now use optional "Add Field" collection
• Changed: multi-text synthesis base params display names prefixed with "ベース" for clarity
• Removed: sampling rate parameter (uses API default)

0.4.5

• Changed: multi-text synthesis base audio params (speed, pitch, intonation, volume, silence) now use optional "Add Field" collection instead of always-visible fields
• Changed: per-text speaker ID moved into optional overrides collection (removed -1 sentinel)
• Changed: JSON input mode now supports all per-text override fields
• Improved: parameter priority for multi-text synthesis: per-text > base collection > API default

0.4.4

• Added: per-text AudioQuery parameter overrides (speed, pitch, intonation, volume, silence) for multi-text synthesis

0.4.3

• Added: multi-text synthesis operation (combine multiple texts into one audio via /connect_waves)
• Added: per-text speaker ID and silence overrides for multi-text synthesis
• Added: GUI and JSON input modes for multi-text synthesis
• Fixed: binary property name now correctly sets the output key (not filename)

0.4.2

• Added: AudioQuery parameters (speed, pitch, intonation, volume, silence, sampling rate, stereo) to simple synthesis
• Updated: official AivisSpeech icon

0.4.1

• Changed: speech synthesis now outputs binary data instead of saving files to disk
• Removed: outputDir parameter (no longer needed)

0.4.0

• Refactored: split monolithic node into modular file structure for better maintainability

0.3.1

• Added: user dictionary operations (Get / Add / Update / Delete)
• Improved: UI labels and descriptions in Japanese

0.1.0

• Initial release
• Speech synthesis operations (Get Speakers, Synthesize, AudioQuery, Synthesize from AudioQuery)

License

MIT