Replicate

Replicate

Developer Tools
Replicate (Cloudflare)
PaidAPI

About

The pioneer of open-source model hosting: run thousands of community and official models with one line of API, billed by the second so you only pay for what you use. Package and deploy your own models with Cog

Share this tool

Our Verdict

Recommended

The model playground that grew up — and got Cloudflare as a parent

Replicate's core promise has not changed since day one: any interesting model, one line of code, pay for the seconds you use. What has changed is everything around it. The Cloudflare acquisition — announced in late 2025 and completed in early 2026 — answered the only serious objection to building on Replicate, namely whether a venture-backed startup would still exist in five years. It will, and its catalog is being wired into Workers AI and one of the largest edge networks on earth. The product itself remains the best browsing experience in AI: tens of thousands of community models to test in the browser, official endpoints for FLUX, Veo-class video and frontier LLMs, and Cog quietly standardizing how models get packaged. The rough edges are the familiar ones — cold starts sting on rarely-used models, community model quality is a lottery, and per-second billing takes discipline to forecast. For dedicated LLM serving at scale, Together and Fireworks are sharper tools. But as the place where you discover, compare and ship generative media and long-tail models, Replicate has no real equal — and for the first time, no expiration risk either.

Best for

  • Product teams shipping image, video or audio features without ML infrastructure
  • Developers exploring and comparing new models the week they trend
  • Researchers publishing models as APIs with Cog

Consider alternatives if

  • You serve LLMs at sustained scale and latency matters (→ Together AI / Fireworks AI)
  • You want the fastest diffusion inference and realtime media APIs (→ Fal.ai)

Supported Platforms

Web AppAPI

Available platforms include Web App and API.

Key Features

Tens of thousands of community models runnable with one line of API code
Official models — FLUX, Seedream, Veo, Claude-class endpoints — priced per output
Cog: open-source tool to package and publish your own models
Per-second hardware billing, from CPU to 8x H100 clusters
Automatic scale-to-zero — you never pay for idle models
Now part of Cloudflare, integrating with Workers AI and the edge network

Pricing

free
No standing free tier, but scale-to-zero billing means trying a model costs cents — you pay only for the seconds it actually runs.
paid
Community models bill hardware per second (CPU $0.36/hr, L40S $3.51/hr, H100 $5.49/hr, up to 8x H100 clusters); official models are priced per output — e.g. FLUX images from about $0.003 each or language models per token. (Verified against the official pricing page, 2026-07-28.)

Use Cases

Adding image, video or audio generation to a product in an afternoon
Prototyping across dozens of models before committing to one
Publishing research models as production APIs via Cog
Running occasional or bursty inference without paying for idle GPUs

Pros

Unmatched model variety — if it made news on GitHub, it is probably here
Scale-to-zero billing is genuinely fair for spiky workloads
Cog has become a de facto standard for packaging ML models
Cloudflare's backing removes the startup-longevity question entirely

Cons

Cold starts on scale-to-zero models add seconds of first-request latency
Community model quality and maintenance vary wildly
Per-second hardware pricing is harder to forecast than per-token plans
Post-acquisition roadmap may gradually favor Cloudflare's ecosystem

Latest Update

2026: Cloudflare completed its acquisition of Replicate — announced late 2025, closed in early 2026 — keeping the brand and API independent while wiring the model catalog into Workers AI. Official endpoints for FLUX, Seedream, Veo-class video and frontier language models keep expanding, with per-second billing and scale-to-zero unchanged.

Subscribe to AI Updates

Get the latest AI tool recommendations, industry insights, and analysis delivered to your inbox.

We respect your privacy. Unsubscribe at any time.