How much RAM does Twenty need?

Twenty requires at least 2 GB RAM for the full stack (server + worker + PostgreSQL + Redis). 4 GB is recommended for production to keep PostgreSQL comfortable under concurrent queries. The server itself idles around 300–400 MB.

Yes — Twenty is fully open source under AGPL-3.0. Self-hosting on your VPS is free; you only pay for the server. There is no per-seat fee, no usage billing and no features locked behind a paid plan on the self-hosted edition.

How does Twenty compare to HubSpot or Salesforce?

HubSpot and Salesforce are cloud-hosted SaaS products with per-seat billing and data stored on their infrastructure. Twenty is self-hosted: your data lives on your server, you pay a flat VPS cost, and you can extend the schema or the API without vendor approval.

Does Twenty have an API?

Yes — every object in Twenty is accessible via a GraphQL API and a REST API. You can query contacts, companies, deals and any custom object, or push data from external tools via webhooks. The API is the primary integration surface, not a bolt-on afterthought.

Can I migrate my HubSpot or Pipedrive data to Twenty?

Yes. Twenty supports CSV import from the workspace settings. Export your contacts and companies from HubSpot or Pipedrive as a CSV, then import them into the matching Twenty object. More complex data migrations (deals with activities) may require a short API script.

Gérer une activité commerciale et administrative.

Twenty

Self-hosted CRM built for developer teams — open-source Salesforce alternative with a modern API and AI-native features.

2 Go (4 Go recommandé) RAM 2 vCPU Port 3000 Disponible

Code source

Stack technique

DockerNode.jsPostgreSQL 16Redis 7

RAM minimum2 Go (4 Go recommandé)

CPU minimum2 vCPU

Port par défaut3000

OS compatiblesubuntu-24.04

Twenty is an open-source CRM (AGPL-3.0) designed from the ground up as a Salesforce and HubSpot alternative that developers can own, extend and self-host. Instead of rigid, opaque schemas, Twenty gives you fully customisable objects, fields and relationships — you model your CRM exactly the way your business works, not the other way around.

Deployed on your ServOrbit VPS via Docker Compose, Twenty runs a Node.js server backed by PostgreSQL 16 and Redis. A background worker handles automations, notifications and scheduled jobs. The REST and GraphQL APIs are first-class citizens: every object, view and relationship is queryable from day one, so integrations and automations are built with real API calls, not screen-scraping workarounds. As of v2.1.0 (April 2026), Twenty is marked production-ready for single-tenant self-hosted deployments.

Fonctionnalités clés

Fully custom objects and fields — model contacts, companies, deals or any entity your business needs

Kanban, list and table views with real-time updates across your team

Built-in notes, tasks and activity timeline linked to any record

GraphQL + REST API covering every object — integrate with n8n, Activepieces or any tool

AI-native: MCP server support, AI field suggestions, semantic search (roadmap)

Background worker for automations, webhooks and scheduled workflows

AGPL-3.0 — no vendor lock-in, no per-seat cloud billing

Quand utiliser cette solution ?

1

Sales pipeline for tech teams

Track leads, deals and customers in a CRM you control — with custom stages, fields and automations that match your real sales process, not a generic template.

2

API-first integrations

Connect Twenty's GraphQL API to your own apps, n8n workflows or Activepieces automations. Sync data bidirectionally without middleware or per-call billing.

3

Data-sovereign CRM for regulated industries

Keep all customer data on your own infrastructure to meet GDPR, local data-residency or confidentiality requirements. No third-party cloud stores your contacts.

Déployer Twenty sur votre VPS

Guide optimisé pour les VPS Cloud ServOrbit.

01

Create the VPS

Order a ServOrbit VPS with at least 2 GB RAM (4 GB recommended for production). Ubuntu 22.04 with Docker pre-installed is the fastest path to a running instance.

02

Set environment variables

Create a .env file with APP_SECRET (a random 32-character string), DB_NAME, DB_USER and DB_PASSWORD for PostgreSQL. These are the only required secrets — Twenty generates its own encryption keys from APP_SECRET.

03

Start the stack

Run docker compose up -d. The stack brings up PostgreSQL 16, Redis 7, the Twenty server (port 3000) and the background worker. On first boot, Twenty runs database migrations automatically — allow 60–90 seconds before the UI is available.

04

Create your workspace

Open http://:3000 and complete the setup wizard: create your admin account, name your workspace and invite team members. The onboarding takes under 5 minutes.

05

Put Twenty behind HTTPS

Point a subdomain (crm.yourdomain.com) to your VPS and install Caddy: apt install -y caddy. Create /etc/caddy/Caddyfile with crm.yourdomain.com { reverse_proxy localhost:3000 }. Set SERVER_URL=https://crm.yourdomain.com in your .env and restart Twenty — Caddy handles TLS automatically.

06

Connect your tools

Use Twenty's built-in API settings to grab your GraphQL endpoint and API key, then wire up n8n, Activepieces or any webhook consumer. Import existing contacts from a CSV in two clicks from the workspace settings.

Questions fréquentes

Twenty is an open-source CRM (AGPL-3.0) built as a self-hosted alternative to Salesforce and HubSpot. It features customisable objects, a modern UI, a GraphQL and REST API, and a background worker for automations. As of v2.1.0 it is production-ready for single-tenant self-hosted deployments.

Solutions complémentaires

Automatisation & Workflows

Activez Twenty sur votre infrastructure.

VPS Cloud dédié — IPv4 incluse, datacenter européen, support inclus. Vos données ne quittent jamais votre serveur.

Configuration recommandée : 2 Go (4 Go recommandé) RAM · 2 vCPU