Replicate review

Replicate is the easiest way to call a wide range of generative models through one API, especially for images, video and audio. Thousands of community and official models are available, and its home page shows runs counted in the millions for models such as gpt-image-2 and nano-banana-2. Most public models bill by the time they run on hardware, and some bill per input and output, for example $0.04 per image for FLUX 1.1 Pro. Each model page carries a cost estimate. You can deploy your own model by packaging it with the open source Cog tool. The cost catch is private models: most run on dedicated hardware, so you pay for setup and idle time as well as active requests, except for fast booting fine-tunes. Hardware runs from a CPU at $0.09 an hour to an H100 at $0.001525 per second, which is $5.49 an hour. For cheap LLM tokens the leaders on this list are better value.

Who it is for

Developers adding image, video and audio generation who want a large public model library and simple per-run billing.

Price

Pay as you go; public models billed per second of hardware or per output; Nvidia H100 at $0.001525 per second ($5.49 an hour).

$0.001525/sec
Quoted from the vendor pricing page on November 7, 2026.

Key features

• Thousands of community and official models callable with one line of code
• Public models billed per second of hardware or per input and output
• Custom model deployment with the open source Cog tool
• Fast booting fine-tunes billed only while active
• Hardware from small CPUs to 8x H100 and H200

Strengths and weaknesses

• Strength: The largest library of image, video and audio models here
• Strength: Per-model cost estimates on each model page
• Strength: Fast booting fine-tunes avoid paying for idle time
• Weakness: Private models on dedicated hardware bill for setup and idle time
• Weakness: H100 at $5.49 an hour is above DeepInfra and Hugging Face
• Weakness: Less focused on low-cost LLM token serving than the leaders

Key facts

• Nvidia H100: $0.001525 per second, $5.49 an hour
• Nvidia T4: $0.000225 per second, $0.81 an hour
• Packaging tool: Cog, open source
• Multi-GPU: Up to 8x H100 or H200

Where it ranks

Replicate is No. 9 in our Best AI Inference Platforms verdict.