This is an n8n community node that integrates Twenty CRM with fully dynamic resource and field discovery. It automatically adapts to your Twenty instance schema, including custom objects and fields, without requiring node updates.
n8n is a fair-code licensed workflow automation platform.
Twenty CRM is a modern, open-source CRM system built for self hosting
Table of Contents:
Features • Installation • Operations • Credentials • Development Status • Bug Reporting • Resources • Credits
🗄️ Supports Custom Databases and Fields: Automatically fetches all standard and custom objects from your Twenty instance
🔓 Access System Databases: View and edit system databases not normally accessible through the Twenty UI (Company/Person Attachments (attachments), Synced Email Metadata (messages), and more)
⚡ Full CRUD + Bulk Operations: Create, Read, Update, Delete, Upsert - all with bulk variants
🧩 Support for Complex Field Types: Template-based inputs for FullName, Links, Currency, Address, Emails, Phones
📋 SELECT/MULTI_SELECT Fields: Dynamic dropdowns with real-time option loading
💾 Smart Caching: Automatic schema caching - fresh on execution, cached in editor for speed
🔄 Dual-API Architecture: Utilizes Twenty Metadata API and GraphQL introspection for complete field coverage for standard and custom fields. REST API used for execution of queries
Modelled after the official Notion N8N node. Unlike traditional n8n nodes with static operations, this node dynamically discovers the Twenty CRM schema at runtime. It queries the Twenty REST and GraphQL APIs to automatically adapt to:
- Support all standard and system Twenty objects (Company, Person, Opportunity, etc.)
- Custom databases and fields created in your Twenty instance
- Schema changes and updates without requiring node updates
Key Architecture:
- Dynamic Schema Discovery: Queries
/metadataendpoint to get available resources and fields - Hybrid GraphQL/REST: GraphQL for mutations, REST API for efficient data retrieval and node queries
- Runtime Query Construction: Builds queries dynamically based on user selections
- Intelligent Caching: Fresh schema on execution, cached in editor UI for performance
- Native n8n Integration: Uses
this.helpers.httpRequestWithAuthenticationfor all API calls
Twenty CRM is an open-source CRM under rapid development. This node stays compatible through dynamic adaptation rather than static operation definitions.
Follow the installation guide in the n8n community nodes documentation.
Quick Install in n8n:
Settings → Community Nodes → Install → n8n-nodes-twenty-dynamic
Or via npm:
npm install n8n-nodes-twenty-dynamicGenerate an API key in Twenty by following the Twenty API documentation.
Quick Steps:
- Open your Twenty instance
- Navigate to Settings → Developers → API Keys
- Click Create API Key
- Copy the generated key
In n8n:
- Click Add Credential and search for "Twenty API"
- Provide:
- API Key: Your Twenty API key (from above)
- Twenty Domain: Your Twenty instance URL (e.g.,
http://localhost:3000orhttps://twenty.example.com)
Important: Use the root domain only, not the GraphQL endpoint:
- ✅ Correct:
https://twenty.example.com - ❌ Wrong:
https://twenty.example.com/graphql
This node dynamically discovers available objects from your Twenty CRM instance and supports the following operations:
- Create: Create a new record with intelligent field inputs
- Get: Retrieve a single record by ID
- Update: Update an existing record (partial updates supported)
- Delete: Delete a record by ID (permanent - cannot be undone)
- Create or Update (Upsert): Smart upsert - create if not found, update if exists (match by ID or unique field)
- List/Search: Retrieve multiple records with pagination (up to 100 records)
Process multiple records in parallel for maximum performance:
- Create Many: Bulk create multiple records (10-20x faster than sequential)
- Get Many: Retrieve multiple records by IDs in parallel
- Update Many: Bulk update multiple records with different field values
- Delete Many: Bulk delete multiple records by IDs
- Create or Update Many (Upsert Many): Bulk smart upsert - create or update multiple records based on unique field matching
Bulk Operations Benefits:
- ⚡ 10-20x faster than sequential operations
- 🛡️ Resilient: Individual failures don't stop the entire batch
- 📊 Detailed results: Each item returns success/error status with index tracking
- 🔄 Parallel execution: Uses
Promise.allSettled()for concurrent processing
- Database Group: Filter databases by type
- All Databases: Show all available objects (default)
- Standard Databases: Main user-facing Twenty objects (Company, Person, Opportunity, Task, Note, etc.)
- System Databases: Internal meta-objects (Views, Filters, Attachments, etc.) - Advanced users only
- Custom Databases: User-created custom objects
- Database: Select the specific object/database to work with (dynamically filtered based on Database Group)
The node automatically provides appropriate inputs based on field types:
- FullName fields (Person.name): Individual First Name and Last Name inputs
- Links fields (domainName, linkedinLink): URL and Label inputs
- Currency fields (annualRecurringRevenue): Amount and Currency Code inputs with dropdown
- Address fields: Street, City, State, Postal Code, Country, and Coordinates inputs
- Emails/Phones: Multiple entry support with primary designation
- SELECT/MULTI_SELECT: Dynamic dropdowns with options from your Twenty instance
- Simple fields: Standard text, number, date, and boolean inputs
- Resource-aware: Same field name behaves differently based on object type (e.g., Company.name is text, Person.name is FullName)
Standard Databases:
- Companies, People, Opportunities, Tasks, Notes
- Workflows, Workflow Runs, Workflow Versions
System Databases:
- Attachments, Calendar Events, Messages, Message Channels, Note Targets, etc.
Custom Databases:
- ✅ All custom databases you create in Twenty
- ✅ Custom fields on standard objects
- ✅ Most custom field types supported
⚠️ Some complex object-based custom fields are work-in-progress
Note: The majority of custom fields are fully supported. If you encounter issues with specific custom field types, please report them on GitHub.
Please report bugs and request features on GitHub Issues.
When reporting bugs, please include:
- Your Twenty CRM version
- Your n8n version
- The database/object you're working with
- Steps to reproduce the issue
- Expected vs actual behavior
- Any error messages
- ✅ Dual-source architecture: Metadata API + GraphQL introspection
- ✅ Complete field coverage: Custom SELECTs + built-in enums
- ✅ All CRUD operations: Create, Read, Update, Delete, List, Upsert
- ✅ Bulk operations: Create Many, Get Many, Update Many, Delete Many, Upsert Many
- ✅ Complex field types: FullName, Links, Currency, Address, Emails, Phones
- ✅ SELECT/MULTI_SELECT: Dynamic options with real-time loading
- ✅ Automatic field type detection: Template-based inputs (no JSON required)
- ✅ Smart caching: Fresh on execution, cached in editor
- ✅ Zero external dependencies: Native n8n helpers only
- ⏳ Advanced filter UI improvements
- ⏳ Add support for remaining Twenty complex field types (Rating, etc.)
- ⏳ Support for Twenty "Views" and "Relations"
- ⏳ Support for get Database and Notion style resources
- 📚 n8n Community Nodes Documentation
- 🔧 Twenty Developer Documentation
- 🔗 Twenty GraphQL API Documentation
- 📝 Changelog - Version history and release notes
- 💻 GitHub Repository - Source code and issues
- 📦 npm Package - Package details
Compatible and tested with Twenty v1.7.6 and n8n Version 1.113.3
Primary Development:
- s-yhc - Dynamic node architecture and custom objects integration
Maintainer:
Community Contributors:
- Testing and feedback from the n8n and Twenty communities
- Bug reports and feature requests via GitHub Issues
License: MIT
Support: GitHub Issues
Latest Version: Check npm or GitHub Releases