Twenty bets open source can beat Salesforce; 56k GitHub stars, $5M seed

An open-source CRM (AGPLv3) in TypeScript/React/NestJS: define custom objects and fields, view data in lists or Kanban, extend via GraphQL/REST APIs, self-host or use the hosted cloud — 'the open alternative to Salesforce, designed for AI'.

That open source wins the biggest software category the way it won databases: companies want to own and customize their CRM; community extensibility beats packaged SaaS.

$5M across two small rounds from ~50 investors; Nov 2024 seed led by Runa Capital with Front founder Mathilde Collin, HubSpot founder Dharmesh Shah, Strapi founders, Pipedrive's former CTO, plus Y Combinator and Automattic.

Malfait, Bochet and Colas des Francs had worked together at Luckey, a mini-CRM for Airbnb guests and hosts that Airbnb acquired. They saw the wave of open-source startups replicating SaaS products — Baserow, Documenso, Formbricks — and realized CRM, the biggest software market, had no modern open-source leader since SugarCRM closed its Community Edition. They launched Twenty in YC S23; the Launch HN post on 2023-07-19 got 415 points and 311 comments and sent the repo to #1 trending on GitHub.

Twenty grew through GitHub instead of enterprise sales: 20,000 stars and more than 300 contributors in the year before Nov 2024, when TechCrunch covered it and it raised a $5M seed led by Runa Capital with ~50 investors, including executives from the very incumbents it targets (Front's Mathilde Collin, HubSpot's Dharmesh Shah, Pipedrive's former CTO). The product is a flexible platform — custom objects and fields, list and Kanban views, GraphQL and REST APIs — built for companies that treat their CRM as the main customer-data repository. A second HN post about twenty.com drew 378 points in June 2024.

Still live and scaling: ~56k GitHub stars, 8.9k forks and 723 contributors as of 2026-09-01, repeatedly #1 trending TypeScript repo on GitHub; open-core model (AGPLv3 code plus paid cloud) intact.

Every dev-tools and SaaS category was getting an open-source leader (Airtable→Baserow, DocuSign→Documenso, Qualtrics→Formbricks); CRM, the biggest market, had none.

CRM data is the company's main customer repository and products get built on top of it, so ownership, extensibility and escape hatches matter more there than in any other business app.

Launching on GitHub/HN instead of through marketing put the product in front of the technical buyers who would later build the integration ecosystem; the repo hit #1 trending on launch day.

The investor list was the thesis in miniature: executives from HubSpot, Front and Pipedrive backed the challenger to their own category.

In markets driven by ecosystems, open source is the wedge: launch the repo, let stars and contributors prove demand, and let community integrations answer the incumbent's platform lock-in.

As of 2026-09-01 Twenty is live and scaling. star-history.com shows ~56k stars, 8.9k forks and 723 contributors, with the repo repeatedly ranked #1 trending TypeScript project on GitHub (Jun 2025, and again in 2026), and a 2026 HN post from a senior engineer confirms the hosted product runs multi-tenant Postgres at scale. The company still runs an open-core model — AGPLv3 code plus a paid cloud at twenty.com — and the original bet is intact: a community-built integration and extension ecosystem as the long-term answer to Salesforce's platform lock-in.