Replicate

Replicate

Replicate is an AI model provider.Tokenando tracks 5 Replicate models, with input pricing from $0.004/M and an average blended cost of $3.373/M. Its flagship model is Llama 3.1 405B (Rep).

Run any open-source model behind a uniform API. Strongest catalogue for image, video and audio generation models.

Founded 2019HQ San Francisco, USAWebsite ↗Official pricing ↗API docs ↗
MODELS TRACKED
5
2 categories
FLAGSHIP
Llama 3.1 405B (Rep)
Live API
MIN INPUT
$0.004/M
cheapest model in family
AVG BLENDED
$3.373/M
across 3 priced models
MAX CONTEXT
128,000
largest window in family

Frequently Asked Questions

How many models does Replicate offer?

Tokenando tracks 5 Replicate models.

How much do Replicate models cost?

Replicate model input pricing starts at $0.004 per million tokens, with an average blended cost of $3.373 per million across the 3 priced models we track.

What is Replicate's flagship model?

Replicate's flagship model is Llama 3.1 405B (Rep). It is the highest-tier Replicate model we track, with input pricing of $9.500 per million tokens.

What model categories does Replicate cover?

Replicate covers 2 categories: frontier and efficient.

Terms used on this page