Replicate
Pay-as-you-go API to run and host open AI models in the cloud without managing infrastructure.
At a Glance
Replicate is an AI tool for Agent & RAG frameworks (category: Development & Code). Free access: Free plan. Documented: 3 plans with prices and limits.
- Category
- Agent & RAG frameworks
- Paid from
- pay as you go / depending on the model (Pay-as-you-go (public models))
- Free Option
- Free plan
- Plans
- 3
- Sources
- 3 verified sources
Category: Development & Code · Agent & RAG frameworks
Features
Replicate is an API (application programming interface) to run open models in the cloud or publish your own with Cog. Billing has no fixed subscription: public model price or instance runtime. GPU means graphics processing unit.
Private instances on dedicated hardware bill while online. Multi-GPU setups such as 8x H100 need committed spend. Enterprise raises GPU limits.
- Run public AI models in the cloud without owning GPUs
- Build and host your own models with Cog
- Access 50,000+ hosted models
- Pay as you go by model price or hardware runtime
- Dedicated instances bill online time including idle; fast-booting fine-tunes bill active time only
Prices and plans
| Tier | Price | Period | What you get |
|---|---|---|---|
| Pay-as-you-go (public models) | pay as you go | depending on the model | Access to 50,000+ hosted models |
| Pay-as-you-go (own models) | $0.000025 | Hardware time | Deploy custom models via Cog |
| Enterprise | contact sales | Upon request | Dedicated account management Higher GPU limits, SLAs |
Prices per vendor, excl. taxes/VAT where applicable.
Limits
No fixed subscription. Public models e.g. FLUX $0.025-$0.04/image, DeepSeek R1 $3.75/million input tokens, video models $0.09-$0.25/sec. Private models billed by hardware uptime: CPU from $0.000025/sec, Nvidia T4 $0.000225/sec ($0.81/hr), Nvidia A100 80GB $0.0014/sec ($5.04/hr), Nvidia H100 $0.001525/sec ($5.49/hr), 8x H100 $0.0122/sec ($43.92/hr). No public free tier. Enterprise on request.
Alternatives and similar tools to Replicate
FAQ
Is Replicate available for free?
Yes. A permanently free plan is published, not a time-limited trial.
Where is the vendor of Replicate based?
Replicate, Inc. (seit 2025/26 Teil von Cloudflare, Inc.) is based in the USA. Regardless of that location, data processing and privacy terms are governed by the vendor's own contract.
How is Replicate billed?
No fixed-price subscription. Public models, e.g. FLUX USD 0.025-0.04/image, DeepSeek R1 USD 3.75/million input tokens, video models USD 0.09-0.25/sec. Your own models are billed by hardware runtime: CPU from USD 0.000025/sec., Nvidia T4 USD 0.000225/sec. (USD 0.81/hour), Nvidia A100 80GB USD 0.0014/sec. (USD 5.04/hour), Nvidia H100 USD 0.001525/sec. (USD 5.49/hour), 8x H100 USD 0.0122/sec. (USD 43.92/hour). No public free allowance. Enterprise on request.
What plans does Replicate offer?
The catalog documents 3 plans with price, billing period, and included features.
Is Replicate currently available?
Replicate is still active: Active.
How we compare
We compare free access, paid entry, plan limits, features and the listed official sources. Unsourced figures stay marked as not published.
Only confirmed entries appear in the catalog. Prices, plans and limits come from listed official vendor pages. Missing figures stay marked as not published. Search results are ordered by Task match by default; category and task lists are alphabetical. Both can be switched to Free access first; neither is a quality ranking. The vendor page always wins.
Edited and checked by the AI Compare Ninja team
Sources
All details regarding plans, limits, and features are based on official vendor documentation and the primary sources verified below.
- https://replicate.com
- https://replicate.com/pricing
- https://www.cloudflare.com/press/press-releases/2025/cloudflare-to-acquire-replicate-to-build-the-most-seamless-ai-cloud-for-developers/
Visit site (opens in a new tab)
Spotted an outdated figure? Vendor pages change without notice. Report an error