Skip to navigation
Getting Started

API Pricing

Chat Completions is billed per token at each model’s rate, with no platform fee or markup. Where a model supports prompt caching, repeated prompt prefixes are cached automatically and billed at the model’s cached-input rate.

Current rates

Rates are set per model and change over time. The model catalog on developer.reka.ai lists the current input, output and cached-input price for every model.

Reading prices from the API

GET /v1/models returns each model’s rates in its pricing object. Values are US dollars per token, as strings.

FieldBilled for
promptEach input token
completionEach output token
input_cache_readEach input token served from the prompt cache

For example, a prompt value of "0.00000011" is $0.11 per million input tokens.

Paying

Add credits before making requests, and track spend on the Usage page. For volume or committed-use pricing, contact contact@reka.ai.

Reka Vision

Reka Vision has its own pricing tiers.