Developers who want to call many image, video and audio models without running GPUs; pay as you go, and costs can climb on H100 time.

Kenning's take

What it does

Replicate lets developers call image, video, speech, music and language models through one API with a few lines of code. It hosts thousands of community open-source models and official models from companies such as Google, OpenAI, Black Forest Labs and ByteDance, and developers can fine-tune models or deploy their own with the open-source Cog tool. Billing is pay as you go, either per second of hardware time or per input and output.

• One API for thousands of public models
• Official models from major labs
• Fine-tuning and custom model deployment with Cog
• Autoscaling dedicated hardware for private models
• Playground to compare models
• Node, Python and HTTP clients

Who it suits. Developers adding AI image, video, audio or text features to their apps without running their own GPUs.

Key facts

Category

Developer tools

Made by

Replicate

How to get it

sign up at replicate.com and call the API; pay for what you use

Works on

API, Web

API price

per model, e.g. FLUX 1.1 Pro $0.04 per image, DeepSeek R1 $3.75 per million input tokens; hardware from $0.000025 a second (CPU) to $0.001525 a second (H100)

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.