TypeSafe AI's Jev now available on AI Gateway
TypeSafe AI's Jev, a probabilistic decision model for software, is now available on AI Gateway. Jev evaluates questions in parallel and returns typed answers with probabilities directly, enabling faster and cheaper decision-making compared to traditional language models for structured reasoning tasks.
is now available on AI Gateway.Jev from TypeSafe AI
Jev is a probabilistic decision model for software: state goes in, typed Choice, Score, and Boolean answers come out.
Regular language models generate text one token at a time, which the application then parses and validates. Jev evaluates all declared questions in parallel and returns typed answers plus probabilities directly. That removes unnecessary text generation and makes it straightforward to automate clear cases while routing uncertain ones to review.
TypeSafe reports Jev was up to 193.6x faster and 444.6x cheaper than LLMs on its workflow evaluations. Example use cases include:
AI SDK 7 exposes Jev through the experimental . Choice selects an option, Score grades an ordered rubric, and Boolean estimates the probability of . Install the current AI SDK (AI SDK 7.0.105 onwards supports the API):evaluatetrueevaluateAPI
Each evaluation specifies:
Call the model with . This example turns one support case into a queue, priority, and refund-review decision, with uncertain routing sent for manual review:typesafe-ai/jev
The result preserves question IDs and Choice keys. TypeSafe reports separate Choice and Score confidence in . Calibrate probabilities and confidence against labeled examples from your workflow.result.providerMetadata.typesafe.confidence
Jev supports and , enabled per request in the example. Evaluation calls also appear in and , count toward , and accept other Gateway provider options in the same object.Zero Data RetentionNo Traininglogscustom reportingbudgetsproviderOptions.gateway
Read the documentation on on AI Gateway for more details.evaluation models
Choosing the next tool or subagent in an agent loop
Deciding whether to continue, retry, ask the user, or stop
Scoring urgency or risk before an action
Verifying model outputs and enforcing guardrails.
: the evaluation model to call,
model: the shared string, object, or array to evaluate, and
state: a map of named decisions to make about that state.
questions
Source: original entry ↗