AI BriefingTypeSafe AI's Jev 1.14 decision model available on AI Gateway
Key point
The new version adds image input support and an experimental long-context mode for processing documents and logs.
Details
TypeSafe AI's Jev 1.14 is now available on AI Gateway. Jev is a decision model for software that accepts state and questions, returning typed Choice, Score, and Boolean answers with probabilities instead of free-form text. It processes all questions in parallel within a single request, supporting classification, routing, scoring, and automated verification.
New Capabilities
The update introduces two key features:
- Image Input: Users can pass images as file parts alongside text in the state.
- Long-Context Mode: An experimental mode (enabled via
providerOptions.typesafe.longContext: true) handles documents, logs, or conversation histories that exceed standard request limits.
Integration and Compliance
The model is accessible using the identifier typesafe-ai/jev-1.14.0. Jev supports Zero Data Retention and No Training options, which can be enabled per request. Decision calls are integrated into Gateway features, appearing in logs and custom reporting, counting toward budgets, and accepting standard provider options.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
TypeSafe AI Raises $870M Series A at $7.5B Valuation Led by Andreessen Horowitz