Model API marketplace

What is Replicate?

Replicate lets developers run, fine-tune, and deploy open-source AI models through hosted APIs.

Category
Model API marketplace
Headquarters
San Francisco, CA
Founded
2019
Employees
Not publicly disclosed
Total funding
Reported ~$58M
Valuation
Not publicly disclosed

What is Replicate?

Replicate is a model api marketplace headquartered in San Francisco, CA. Replicate lets developers run, fine-tune, and deploy open-source AI models through hosted APIs.

Replicate operates in model api marketplace and is best understood by its product surface: Hosted model APIs, Cog, Fine-tuning, Deployments, Model marketplace. Customers are developers, startups, and teams that want model APIs without managing infrastructure. Revenue is not disclosed; monetization is developer usage of hosted models and private deployments.

The company was founded in 2019, is headquartered in San Francisco, CA, and has not publicly disclosed employees according to public reporting or company-visible signals. Its market position is shaped by the acceleration of AI infrastructure and application spend, but most private-company financials remain reported estimates rather than audited disclosures.

For buyers and sellers, Replicate is a useful account because the company sits close to production AI budgets. The relevant buying centers are usually engineering, research, infrastructure, product, security, and enterprise operations rather than a single generic AI owner.

What does Replicate offer?

Replicate's public product surface includes Hosted model APIs, Cog, Fine-tuning, Deployments and related enterprise capabilities.

  • Hosted model APIs· Product / capability
  • Cog· Product / capability
  • Fine-tuning· Product / capability
  • Deployments· Product / capability
  • Model marketplace· Product / capability

How does Replicate make money?

Replicate monetizes through a mix of self-service usage, subscriptions, infrastructure capacity, licensing, and enterprise contracts depending on the product line.

Replicate prices by compute time / hardware usage for public and private model runs, with enterprise options for scale.

Growth is driven by production AI usage: more prompts, model calls, GPU hours, private deployments, seats, storage, or managed workflows depending on the product. Where exact pricing is not published, the profile treats enterprise terms as contract-priced rather than inventing seat prices.

The buyer motion is therefore both developer-led and enterprise-led. Individual builders may start with a free tier, open-source project, or usage-based API, while larger accounts typically involve procurement, security review, data-governance review, and committed capacity or platform agreements.

Who leads Replicate?

Replicate's publicly visible leadership includes Ben Firshman, Andreas Jansson.

  • Ben FirshmanCo-founder & CEOPublic role as of June 2026Leads product and company strategy.
  • Andreas JanssonCo-founderPublic role as of June 2026Co-founder and technical leader.

How do you contact Replicate's leadership?

No verified public executive email pattern was confirmed for Replicate; use the company's official contact, sales, support, or press paths rather than guessed personal emails.

Email formatNo verified public executive email format; use official contact channels.

How much funding has Replicate raised?

Replicate has reported ~$58m, with valuation/status currently described as Not publicly disclosed.

Replicate's disclosed funding history is: 2020: Seed — Early venture backing for model deployment. 2022: Series A — Reported $17.8M Series A. 2023: Series B — Reported $40M Series B led by Andreessen Horowitz.

Because several AI companies use strategic partnerships, cloud credits, debt, project finance, or reported-but-not-yet-filed private rounds, this profile separates disclosed funding from directional reports. The valuation field is reported as Not publicly disclosed; it should be rechecked before financial decisions.

The seller signal is budget maturity. Companies with late-stage or infrastructure-scale financing usually have formal procurement and security processes, while earlier labs and open-source companies may centralize decisions with founders and technical platform leaders.

How did Replicate get here?

Replicate's timeline runs from its 2019 founding to its current model api marketplace position.

  1. 2019FoundedReplicate starts as a model-running platform.
  2. 2020CogThe team releases packaging tooling for ML models.
  3. 2022Series AReplicate raises for hosted model APIs.
  4. 2023Series BThe company raises $40M.
  5. 2024Model marketplace growthReplicate becomes a common path for running image/audio/video models.
  6. 2026Open-model servingThe company competes in a broader inference-platform market.

Who are Replicate's competitors?

Replicate competes with Hugging Face, Together AI, Baseten, Modal, Fireworks AI depending on the buyer's use case.

  • Hugging FaceHugging Face overlaps with Replicate for buyers evaluating model api marketplace alternatives, but differs by product depth, deployment model, geography, or ecosystem.
  • Together AITogether AI overlaps with Replicate for buyers evaluating model api marketplace alternatives, but differs by product depth, deployment model, geography, or ecosystem.
  • BasetenBaseten overlaps with Replicate for buyers evaluating model api marketplace alternatives, but differs by product depth, deployment model, geography, or ecosystem.
  • ModalModal overlaps with Replicate for buyers evaluating model api marketplace alternatives, but differs by product depth, deployment model, geography, or ecosystem.
  • Fireworks AIFireworks AI overlaps with Replicate for buyers evaluating model api marketplace alternatives, but differs by product depth, deployment model, geography, or ecosystem.

Replicate — frequently asked questions

Agent CTA Background

Revenue work. On autopilot.

Start Free TrialBuilt for revenue teams who care about quality.