Awesome Jev: early projects and use cases
A community-curated list gathers early Jev projects and ideas for developers who already have API access.
Painting a Bob Ross landscape with Jev
Kartikey Pandey demonstrates an experiment: the code generates the canvas, and TypeSafe's Jev model decides each paint by answering multiple-choice questions. With 37,856 decisions, it painted a Bob Ross-style landscape in 35 seconds at a cost of 24 cents. Video shown at 4x speed.
An AI that can't write or draw just painted a Bob Ross landscape. Paint-by-numbers: my code makes the sheet, @typesafeai's Jev picks every paint. 37,856 decisions. 35 seconds. 24 cents. Video at 4x. The trick: Jev only answers multiple-choice questions. cc @CompleteSkeptic
· 0 likesOpen on X
More like this
Training an MLP on Qwen 4B to Replicate Jev Behavior
The author shares an open-source project openjev that trains an MLP on qwen 4b to work like Jev, with a Hugging Face link. Link
Swapping an agent's tool-calling reasoning with Jev
The author shares replacing an agent's tool-calling reasoning with Jev, using Fable, Astra, and Opus; early results show drastic savings with very close results, with a demo video promised.
