Replicate
Free tierRun open-source AI models in the cloud via a simple API — images, video, audio, and LLMs.
Free tier available·Technical·Powered by Multi-provider (Black Forest Labs, OpenAI, Google, Anthropic, ByteDance, xAI, Alibaba, Recraft, Minimax, and community models)·API available·Open source
Key strengths
Run thousands of community and official AI models via APIFine-tune models with custom dataDeploy custom models using Cog (open-source packaging tool)Automatic scaling up and down to zeroPay-per-use pricing with no idle charges for public modelsWide model variety: image, video, speech, music, LLMs
Free tier + paid plans
San Francisco, US
Founded 2021
Facts verified against replicate.com on
Recent changes
- Plans updated: Pay-per-use: public models billed by time (e.g. CPU $0.000100/sec, Nvidia T4 $0.000225/sec, Nvidia L40S $0.000975/sec, Nvidia A100 80GB $0.001400/sec) or by input/output (e.g. flux-1.1-pro $0.04/image, deepseek-r1 $0.01/1K output tokens). Private models billed for all online time including idle.
- Models updated: Multi-provider (Black Forest Labs, OpenAI, Google, Anthropic, ByteDance, xAI, Alibaba, Recraft, Minimax, and community models)
Get alerts when Replicate changes
Free. We'll email you when pricing, the free tier, API access or open-source status changes.
No ratings yet
Replicate's API follows a standard request/response pattern with support for asynchronous predictions. Key integration points:
- SDKs: Official libraries for Node.js and Python; all functionality is also accessible over plain HTTP
- Model packaging: Use Cog (open-source) to containerize custom models and push them to Replicate for deployment
- Fine-tuning: Supported for select models using your own training data via the API
- Scaling: Automatic scale-to-zero for public models; private/custom model deployments are billed for all online time, including idle
- Pricing model: Usage-based with no flat fee. Public models are billed by compute time (e.g., CPU $0.000100/sec, Nvidia T4 $0.000225/sec, Nvidia L40S $0.000975/sec, Nvidia A100 80GB $0.001400/sec) or per input/output unit (e.g.,
flux-1.1-proat $0.04/image,deepseek-r1at $0.01/1K output tokens). Minimum rate starts at $0.000025. - Auth: API-key based authentication
- Self-hosting: Not available; compute is fully managed by Replicate
Alternatives
See all alternatives Alternatives to Replicate
Is Replicate yours? Claim the listing & get the badge