Problem

High costs of prompting is one of the challenges in using AI in translation.

A solution to reduce costs is batch compression of prompts.

Solution

Implement batch compression for queries to LLMs.

Add a flag in AI language resource properties, that determines, if the batch will be send compressed or not.

Use the information in this article as inspiration:

https://slator.com/how-to-balance-cost-quality-in-ai-translation-evaluation/