Jev by Jevapi - Turn text into clear, actionable classification, evaluation and score decisions

Core Capabilities

Jev processes any input text to produce structured, machine-readable results that can be used directly in other applications. It supports three distinct operation types for user-defined questions:

1. Multi-choice classification: Assign input text to one of your predefined custom labels

2. Yes/No evaluation: Return a clear boolean result for binary assessment questions

3. Numeric scoring: Assign a ranked numerical value to input content across your custom 2-10 level scale

Supported input text types include customer support messages, news articles, user reviews, and unstructured business data.

Workflow

The standard workflow for a Jev run follows three simple steps:

1. Paste your source text into the input field

2. Define your custom questions, select the operation type, and add required labels or score levels as needed

3. Run the process to return structured choices, probability values, and final scores. All generated outputs are formatted for direct use in connected applications via the public API.

Usage Terms

The platform uses a credit-based usage system:

• Free requests are consumed first, with one free credit applied per successful request

• After free credits are exhausted, paid credits apply at the rate of 1 paid credit for 1 million input tokens Users are advised to review all high-stakes decisions returned by the tool. Submitted content is routed to the model provider configured by the workspace administrator. Outputs from the tool are model-generated estimates and are not guaranteed to be 100% accurate.