Startup profile for Replicate: latest funding, latest known valuation, total disclosed funding, investors, status, sources, and why the company is interesting. Part of the Venture Capital Tracker startup directory.
Venture Capital Tracker logo
Venture Capital Tracker

Startup profile · funding coverage

Replicate funding, valuation and investors

Cloud platform to run and deploy open-source ML models via API.

Keep track of Replicate

Save this profile to your VCT watchlist for a quick return.

View watchlist

Funding, valuation & investors

Answer-first snapshot

Latest funding

Series B

$40M · December 2023

Latest known valuation

$350M

Series B · December 2023

Total disclosed equity funding

$53M

Excludes debt, grants, acquisitions, secondaries, and IPO proceeds.

Current status

private

series b · San Francisco, California, USA

Investors in latest funding

Lead: Andreessen Horowitz

Other: Sequoia

Sources: latest funding Last verified: 2026-07-25

Overview

Replicate lets developers run open-source machine learning models (image, text, audio) via simple API without managing GPUs. Founded by Ben Firshman and Andreas Jansson, it became popular for Stable Diffusion and LLM hosting. Sequoia led $12.5M Series A (Feb 2023); a16z led $40M Series B (Dec 2023) at $350M with NVentures and YC participating.

Why Replicate is interesting

a16z-led $40M Series B (2023) at $350M as 'npm for ML models' — Sequoia led A; NVIDIA NVentures joined as GPU infra strategic.

Product & use cases

Replicate hosts model containers on cloud GPUs; developers call REST API with model version pins — pay per second of inference.

  • Running Stable Diffusion and image models via API
  • Prototyping LLM apps without GPU DevOps
  • Custom model deployment from Cog containers
  • Fine-tuned model hosting for startups

Key facts

  • Series B (Dec 2023): $40M at $350M led by a16z — Forbes
  • Series A (Feb 2023): $12.5M led by Sequoia
  • Cog open-source tool for packaging models
  • Popular during Stable Diffusion boom

Funding history (newest first)

Series B

2023-12 $40M Valuation: $350M

Source: https://replicate.com/blog/series-b

Series A

2023-02 $12.5M

Source: https://replicate.com/blog/series-a

Investors in our directory

Funds linked from Replicate's profile — open a fund page for stage focus and related deal articles.

Andreessen Horowitz Sequoia Capital

Competitive landscape

Edge: Developer experience and model catalog breadth at launch of generative AI app wave.

Model hosting commoditizing as hyperscalers and Hugging Face expand. Replicate competes on DX and Cog packaging format.

  • Hugging Face Inference direct

    Model hub with hosted inference endpoints.

  • Modal / Banana direct

    GPU serverless for ML workloads.

  • AWS SageMaker incumbent

    Enterprise ML hosting.

Notable stories

  • Replicate surged when developers needed one API call to run SD models without owning GPUs — timing perfect for 2023 generative AI explosion.

Industries

AI & Machine Learning Developer Tools Infrastructure & Cloud

FAQs about Replicate

Practical answers founders, operators, and investors typically search for.

Replicate runs open-source ML models in the cloud via API — pay per inference second.
a16z (/fund/andreessen-horowitz) led Series B. Sequoia (/fund/sequoia) led Series A.
Open-source tool by Replicate for packaging ML models as reproducible containers.
HF is model hub + community; Replicate emphasizes dead-simple API deployment.
Ben Firshman and Andreas Jansson.
December 2023 — $40M at $350M valuation.
Yes — many open LLMs and diffusion models available on replicate.com/explore.
Pay per compute second based on GPU type — see replicate.com/pricing.
~$52.5M across Series A and B plus seed.
Replicate removes GPU DevOps; self-hosting cheaper at scale but higher setup cost.
Website All startups Funding news

By Venture Capital Tracker

Last updated:

Editorial note: AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.