Skip to main content
vLLM logo

vLLM

Updated Mar 24, 2026

Everything you need to know about vLLM — from their funding history to who's backing them.

🇺🇸United StatesBootstrappedAI Infrastructure

Out of 100

🇺🇸

HQ

United States

🗓

Founded

2023

🏷️

Stage

Bootstrapped

👥

Team

1-50

vLLM is an open-source high-throughput and memory-efficient inference and serving engine for large language models, developed initially at UC Berkeley and widely adopted in production AI deployments. The project introduced PagedAttention, a novel memory management technique that significantly increases GPU utilization during LLM inference by managing key-value cache memory analogously to how

Is this your company? Claim it →

Key Facts

  • vLLM was founded in 2023 by Woosuk Kwon and is headquartered in United States.
  • vLLM employs approximately 1-50 people.
  • vLLM operates in the AI Infrastructure category and is currently at the Bootstrapped stage.
  • vLLM has an Awaira Score of 45/100, ranking it among the tracked AI companies on Awaira.
  • vLLM's most recent funding round was a Seed of $1.71M closed in January 2023.

Source: Awaira · Updated Mar 24, 2026

N/A

Post-money

N/A

All rounds

45/100

2023

1-50 employees

Strengths

Growing team of 1-50 employees

Based in the US, the world's largest AI market

Considerations

!

Valuation not publicly disclosed

!

Limited disclosed funding

Analysis based on publicly available data. Not investment advice.

⚠️ ESTIMATE
💵Est. Annual Revenue
Revenue not disclosed

Awaira estimate based on public data. Not financial advice.

How does vLLM stack up against competitors?

MetricvLLMCoreWeaveGraphcoreNebius
Awaira Score45/10095/10085/10084/100
Total RaisedN/A$2.38B$767M$2B
ValuationN/A$49B$600M$25B
StageBootstrappedPublicAcquiredPrivate
Founded2023201720162023
Employees1-501800500-10001,000
Country🇺🇸🇺🇸🇬🇧🇳🇱
W

Woosuk Kwon

Creator & Lead

View founder profile →
StageBootstrapped
Employees1-50
Country🇺🇸 United States
Connect
Inactive
1 round
SeedLatest

$1.71M

Jan 2023

Frequently Asked Questions

What is vLLM's valuation?
No public valuation has been disclosed yet. The company operates in the AI Infrastructure space and may reveal this in a future round.
Who invested in vLLM?
The investor base hasn't been publicly disclosed. The company may have institutional or angel backers that remain unnamed.
When did vLLM last raise funding?
A Seed of $1.71M represents the most recent capital raise in January 2023.
How many employees does vLLM have?
Around 1-50 people work there, with headquarters in United States.
What does vLLM do?
vLLM is an open-source high-throughput and memory-efficient inference and serving engine for large language models, developed initially at UC Berkeley and widely adopted in production AI deployments. The project introduced PagedAttention, a novel memory management technique that significantly increases GPU utilization during LLM inference by managing key-value cache memory analogously to how operating systems manage virtual memory pages. The company operates in the AI Infrastructure sector from United States.
Who founded vLLM?
Woosuk Kwon established the company in 2023. It's based in United States. Woosuk Kwon serves as Creator & Lead.
Is vLLM profitable?
Profitability details aren't publicly disclosed. The company has not disclosed its funding and is currently at the Bootstrapped stage.
Where is vLLM headquartered?
Headquarters are in United States 🇺🇸. The company operates in the AI Infrastructure sector and was established in 2023.
⚔️

Compare vLLM

See how vLLM stacks up against competitors in valuation, funding, team size, and Awaira Score.

Browse all comparisons →