A Microsoft Research library that compresses long prompts by dropping low-value tokens, so the model reads less and costs less.
pip install llmlinguaLLMLingua uses a small, well-trained language model to find and remove tokens that do not matter in a prompt. The shorter prompt then goes to your main model. It comes from Microsoft Research papers (EMNLP 2023, ACL 2024).
pip install llmlingua
Pass your prompt through its PromptCompressor before you send it to the model.
rtk-ai · Tool / CLI
A CLI proxy that compresses shell output before your coding agent reads it, so every command costs fewer tokens.
diegosouzapw · Tool / CLI
A free AI gateway with one OpenAI-compatible endpoint that routes across hundreds of providers, many with free tiers.
Alishahryar1 · Tool / CLI
A local proxy that runs Claude Code, Codex and other agents on free-tier providers, falling back automatically when one goes down.