$100K ARRin 4 years

3 milestones

Acquired

Sold to Google DeepMind (acquihire) + Cognition AI (IP)for $2.4Bas of July 2025Source

Pre-acquisition revenue: $82M ARR

Google DeepMind hired Varun Mohan, Douglas Chen, R&D team in $2.4B acquihire July 11, 2025. Cognition AI separately acquired Windsurf's IP, product, brand. Pre-deal ARR ~$82M.

Windsurf: Doble pivote de infraestructura GPU a editor de código IA de $2.4B

Founder

Varun Mohan

Co-Fundadores•Technical•USA

Commitment

Full-time

Experience

First-time

Product

Windsurf

Editor de código nativo de IA (anteriormente Codeium) con IA agéntica que entiende bases de código completas, permitiendo a los desarrolladores construir software más rápido a través de lenguaje natural.

Type

SaaS

Industry

AI / ML

Model

Freemium

Marketing Strategy

How Varun acquired customers

Growth Channel

Boca a Boca

Also Used

ComunidadesCold Outreach

Tech Stack

Tools used to build Windsurf

VS Code (fork)Proprietary AI modelsTypeScriptReactPythonGPT-4o

The Full Story

TL;DR

Varun Mohan mató un negocio rentable de GPU para perseguir código IA. Después de pivotar dos veces, Windsurf creció de $12M a $100M ARR en 4 meses — luego Google lo adquirió por $2.4B.

Varun Mohan y Douglas Chen se conocieron en la secundaria en Sunnyvale, California, estudiaron CS juntos en MIT, y en junio de 2021 fundaron Exafunction — una plataforma de optimización y virtualización de GPU. El producto era rentable. Luego vieron que la IA generativa venía y tomaron una decisión contraintuida: matar el negocio rentable y pivotar.

El Doble Pivote

En 2022, cambiaron el nombre a Codeium e lanzaron una extensión de autocompletado de código potenciada por IA. Era bueno, pero jugaba en un campo abarrotado con GitHub Copilot. La extensión gratuita acumuló millones de usuarios individuales, pero los ingresos fueron modestos — alrededor de $12M ARR a finales de 2024. Necesitaban una apuesta más grande.

El IDE Agéntico

El 13 de noviembre de 2024, lanzaron el Editor Windsurf — posicionándolo como el "primer IDE agéntico" con su tecnología Cascade propietaria para razonamiento de múltiples pasos y Flow para conciencia de código en tiempo real. Dentro de 4 meses, 1 millón de desarrolladores lo estaban usando. Los ingresos crecieron 8x: de $12M a $100M ARR entre finales de 2024 y abril de 2025.

La Salida

En abril de 2025, cambiaron el nombre de Codeium a Windsurf. OpenAI supuestamente ofreció $3B para adquirirlos. Finalmente, Google pagó $2.4B por una licencia de tecnología no exclusiva y talento clave — Varun y Douglas se unieron a Google DeepMind en julio de 2025 con 350+ clientes empresariales incluyendo JPMorgan Chase y Dell.

Fundado: Junio de 2021 (como Exafunction)
Pivotes: 2 (infraestructura GPU → autocompletado → IDE completo)
$12M ARR: Finales de 2024
$100M ARR: Abril de 2025 (8x en 4 meses)
Acuerdo con Google: $2.4B (Julio de 2025)
Clientes empresariales: 350+ (JPMorgan, Dell)

Key Takeaways

Mata tu producto rentable si surge una oportunidad más grande — Exafunction era rentable pero hubiera sido commoditizada

El nivel gratuito construye distribución — regalar extensiones de Codeium a millones creó la base de usuarios para el aumento del IDE

Empresarial es donde está el dinero real — 350+ clientes empresariales impulsaron la mayor parte de los ingresos

Los amigos de primaria pueden construir resultados de múltiples mil millones cuando la confianza es profunda

Cada pivote fue una apuesta en un mercado más grande — infraestructura GPU → autocompletado → IDE agéntico

Originally published onVarun Mohan (X/Twitter)

Founder proof brief

A one-page proof brief turns Varun's Windsurf playbook into your next test.

See what it covers before you sign up: the pattern worth copying, the channel to test, and the risk to avoid when you adapt this story.

Pattern

$100K ARR

Channel

Boca a Boca

Output

Action checklist

What a proof brief covers

A one-page brief that turns Varun's pattern into a testable plan
Channel bet: start from Boca a Boca
A risk check before copying the Windsurf playbook

See what a proof brief covers

See what it covers first. Create an account only if you choose to unlock.

Inspired by Varun's journey? Generate a business idea in the AI / ML space using AI and real founder data.

Milestone Journey

Varun achieved 3 milestones on the path to $100K ARR

Primer Cliente

1 year

October 2022

Avg: 3 months

+1 year to next milestone

$10K MRR

3 years

December 2024

Avg: 1 year

+6 months to next milestone

$100K ARR

4 years

April 2025

Avg: 3 years

4 years

Total journey time

Milestones achieved

Varun's Path to $100K ARRPremium

The journey, decisions, and context behind this milestone

How Varun actually did it

5/7 funnel stages sourced

Reconstructed from Varun's own published sources. Quoted claims are the founder's words, verified verbatim; anything marked as our analysis is exactly that.

Where the idea came from

Pivot from prior productPivot from Exafunction (GPU virtualization infrastructure for deep learning workloads, founded 2021), then Codeium (a free AI coding autocomplete and chat extension, 2022), before becoming the standalone Windsurf Editor IDE (launched November 2024).

Windsurf's November 2024 launch looked sudden from the outside, but Varun has said the team had effectively been building toward it since the company's 2021 founding as Exafunction (a GPU virtualization business) and its 2022 pivot into Codeium, a free AI coding extension.

“Windsurf may only seem like a few months old, but the reality is that we've been building towards this for nearly 4 years.”

Varun Mohan's own LinkedIn post promoting his Lenny's Newsletter conversation

Validation

Built first, validated after

MvpValidated

Codeium, the free coding extension that preceded the Windsurf Editor, had essentially no users through mid-2022 right after the pivot away from Exafunction. Rather than a formal presale, survey, or waitlist, the signal that the pivot was working came from shipping the free extension and watching adoption: it grew to 300,000 active users and 100 enterprise customers by 2023.

Varun described the free extension going from zero users in mid-2022 to 300,000 active users and 100 enterprise customers in 2023 as the moment the pivot's risk paid off.

Per Unite.AI written interview: "Varun Mohan, Co-Founder & CEO of Codeium"

Acquisition channels, and whether they worked

Word of Mouth · Free Codeium extension (VS Code and other IDEs)ScalingWorked

Codeium was made free for individual developers from the start of the post-pivot product, which Varun described as a deliberate accessibility and adoption play that came with real infrastructure cost and security trade-offs to manage.

Per Unite.AI written interview: "Varun Mohan, Co-Founder & CEO of Codeium"

What didn't work

Tech choiceScalingCost: both

In June 2025, Anthropic cut Windsurf's direct API access to its Claude models; several Windsurf customers responded by switching to competitors such as Cursor that retained full Claude access.

“cut Windsurf's direct access to its Claude AI models in June”

TechCrunch: "Cognition, maker of the AI coding agent Devin, acquires Windsurf"

What they did instead: Once Cognition acquired Windsurf's IP and product in July 2025, Windsurf regained full access to Claude AI models.

FundraisingScalingCost: time

OpenAI reportedly agreed to acquire Windsurf for around $3 billion in spring 2025, but the deal never closed and the offer expired.

“OpenAI's $3 billion offer to acquire Windsurf expired, clearing the way for the AI coding startup to explore other options”

TechCrunch: "Cognition, maker of the AI coding agent Devin, acquires Windsurf"

What they did instead: Google reached a separate roughly $2.4B agreement to bring Varun Mohan, Douglas Chen, and part of the R&D team into Google DeepMind, and days later Cognition AI acquired Windsurf's remaining IP, product, brand, and team.

How it changed as they grew

Today: 250 people (Roughly 250 people at the time of the July 2025 Google agreement. Varun Mohan, Douglas Chen, and part of the R&D team departed for Google DeepMind; most of the rest of the team stayed with the Windsurf product, which Cognition AI then acquired.)

Series a plus · $150,000,000 · 2024-08

What broke as they grew

Even after strong revenue growth, Windsurf remained the smaller player next to Cursor, whose annualized recurring revenue had reached roughly $500M by mid-2025 versus Windsurf's reported $82-100M, a relative-scale gap that shaped how acquirers and investors weighed the two companies against each other.

Per TechCrunch: "Cognition, maker of the AI coding agent Devin, acquires Windsurf"

Sources (1 in the founder's own words)

Unite.AI written interview: "Varun Mohan, Co-Founder & CEO of Codeium" · third party · 2024-03
Codeium company blog: "Windsurf Launch" (unsigned, byline "Codeium Team") · third party · 2024-11
Windsurf company blog: "The Next Chapter: Renaming to Windsurf" (unsigned; now hosted on Cognition's devin.ai after Cognition acquired the Windsurf brand/blog) · third party · 2025-04
Windsurf company blog: "The Next Stage of Windsurf" (unsigned, announcing the Google agreement) · third party · 2025-07
Varun Mohan's own LinkedIn post promoting his Lenny's Newsletter conversation · founder-authored · 2025-04
Business Insider: "Windsurf's CEO is headed to Google DeepMind, not to OpenAI" (carries a joint Mohan/Chen statement given to the reporter) · third party · 2025-07
TechCrunch: "Windsurf's CEO goes to Google; OpenAI's acquisition falls apart" · third party · 2025-07
TechCrunch: "Cognition, maker of the AI coding agent Devin, acquires Windsurf" · third party · 2025-07
BusinessWire press release: "Codeium Reaches $1.25B Valuation with $150M Series C Funding Led by General Catalyst" (quotes Varun Mohan) · third party · 2024-08

Persistence

Projects attempted before finding success

failed projects before this one worked

Learned from a previous attempt

Launch Strategy

How they introduced the product to the world

Soft Launch

Initial go-to-market approach

Low risk — lets you iterate quietly

Validation

How they tested demand before building

MVP

Method used to confirm market interest

Most common approach — build and learn fast

Launch Pricing

Price point when the product first launched

Menos de $20/mo

Initial pricing strategy

Starting Audience

Whether they had followers before launch

Existing Audience

Leveraged existing followers

Having an audience accelerates early growth

Time Investment

Average weekly hours during building phase

80 hrs

per week on average

Full-time dedication

Initial Investment

Capital required to get started

$3,000,000

in startup costs

Significant upfront investment

Biggest Challenge

Pivotar dos veces — de optimización de GPU a autocompletado de código a IDE completo — antes de encontrar el producto correcto

Unlock Varun's Full Journey

See the complete breakdown: launch strategy, validation methods, startup costs, expert analysis, replication playbook, and more actionable insights.

Upgrade to Premium

Instant access to all founder journeys

Frequently asked questions

Who acquired Windsurf?

Windsurf was acquired by Google DeepMind (acquihire) + Cognition AI (IP) for $2.4B on July 2025. Pre-acquisition revenue was $82M ARR. Google DeepMind hired Varun Mohan, Douglas Chen, R&D team in $2.4B acquihire July 11, 2025. Cognition AI separately acquired Windsurf's IP, product, brand. Pre-deal ARR ~$82M.

What is Windsurf?

How long did it take Windsurf to reach $100k arr?

Varun Mohan took 4 years to reach $100k arr with Windsurf.

Was Varun Mohan a solo founder?

No. Varun Mohan built Windsurf with co-founders.

What marketing channel did Windsurf use to grow?

Windsurf grew primarily through Boca a Boca.

What industry is Windsurf in?

Windsurf operates in AI / ML.

Share this story:

Explore similar stories

$100K ARR Boca a Boca AI / ML Co-Fundadores

Enjoyed this story?

Get more founder journeys like this delivered to your inbox every week.

Join founders learning from real success stories

Founder stories

Browse Stories

The Full Story

TL;DR

Varun Mohan mató un negocio rentable de GPU para perseguir código IA. Después de pivotar dos veces, Windsurf creció de $12M a $100M ARR en 4 meses — luego Google lo adquirió por $2.4B.