14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis, intelligent content routing. Zero LLM inference cost. MIT licensed.

JavaScript/TypeScript implementation of LLMLingua-2 (Experimental)

Python command-line tool for interacting with AI models through the OpenRouter API/Cloudflare AI Gateway, or local self-hosted Ollama. Optionally support Microsoft LLMLingua prompt token compression

Lossless-first prompt compression for JSON, YAML, CSV, and Markdown. Library, CLI, MCP server, desktop app, and browser extension.

A self-improving knowledge base about LLM agent infrastructure

Rolling context compression for Claude Code — never hit the context wall. Auto-compresses old messages while keeping recent context verbatim. Zero config, zero latency. Works as a Claude Code plugin.

CUTIA: compress prompts while preserving quality

A curated list of strategies, tools, papers, and resources for reducing LLM token costs and improving efficiency in production.

This repository is the official implementation of Generative Context Distillation.

TOON for TYPO3 — a compact, human-readable, and token-efficient data format for AI prompts & LLM contexts. Perfect for ChatGPT, Gemini, Claude, Mistral, and OpenAI integrations (JSON ⇄ TOON).

LLMLingua-2 prompt compression hook for Claude Code — cut token usage by ~55%

Compress LLM Prompts and save 80%+ on GPT-4 in Python

API gateway for LLM prompt compression with policy enforcement built on LLMLingua. Demonstrates cost control, prompt safety, and LLM execution boundaries.

Enhance the performance and cost-efficiency of large-scale Retrieval Augmented Generation (RAG) applications. Learn to integrate vector search with traditional database operations and apply techniques like prefiltering, postfiltering, projection, and prompt compression.

End-to-End Python implementation of CompactPrompt (Choi et al., 2025): a unified pipeline for LLM prompt and data compression. Features modular compression pipeline with dependency-driven phrase pruning, reversible n-gram encoding, K-means quantization, and embedding-based exemplar selection. Achieves 2-4x token reduction while preserving accuracy.

A fast, Unix-style CLI tool for semantic prompt compression. Cuts LLM prompt tokens by 10-20x with >90% fidelity, saving costs and latency.

This repository contains the code and data of the paper titled "FrugalPrompt: Reducing Contextual Overhead in Large Language Models via Token Attribution."

LLM context compression proxy — 40-70% token savings, zero code changes

PirateBao is a TypeScript/Bun agent-skill package for terse pirate-speak AI coding replies that preserve technical detail while cutting filler, with hooks, compressor CLI, OpenCode/Codex/Claude/Gemini cargo, .bao validation, npmjs gates, and token eval checks.

Prompt Cloud Intent Compression