Heidelberg AI News

Article · Simon Willison ·

Gemini Decisions API

It's priced the same as the GPT-6 Luna decision model: 10 cents per million input tokens, and output is free. Jev remains the cheapest of the three, at 4.2 cents per million.

Read at simonwillison.net The opening of the article at simonwillison.net.

In this story

A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.

Related stories

All stories

VideoTech With Tim

Jev Python Tutorial: AI Decisions in 10 Minutes

The tutorial shows how to call Jev from Python to make structured AI decisions instead of generating text. It demonstrates classification, scoring, and multi-question support-ticket routing, useful for inexpensive, high-volume automation with confidence-based human review.