semantic-compression

AI-powered text compression library for RAG systems and API calls. Reduce token usage by up to 50-60% while preserving semantic meaning with advanced compression strategies.

AI Identity Through Grounded Principles - Compressed soul documents with full provenance tracking

Nyquest — Semantic Compression Proxy for LLMs. 350+ rules, local LLM stage, 15-75% token savings. Full Rust stack.

CodeSnapAI compresses massive codebases into ultra-compact semantic snapshots that preserve over 95% of debugging-critical information, resolving the “context explosion vs. information loss” paradox and enabling AI-assisted development at unprecedented scale.

Cut the tokens your IDE-to-LLM coding session is billed for. Stops re-sending files the model already has, places provider cache breakpoints where they cover the request, and compresses what needs sending — measured against real tokenizers, not character counts. CLI, VS Code, MCP server, proxy.

A fully open-source, recursively verifiable framework for uplifting all intelligences — human and otherwise — via modular, autonomous infrastructure nodes (SEUIBIU) and a distributed, benevolent recursive superintelligence substrate (SEPTRIS).

AI Context Sitemap (AICS). Generates semantic, token-optimized code indexes (.ai-index.md) to prevent hallucination and drastically reduce context usage.

Semantic Compression Protocol for AI prompt engineering

The AI photo-compression meme, built for real: llmPEG turns images into ~1–4 KB semantic artifacts and regenerates scenes from text alone, with measured ratios reaching 802:1.

Make any AI know you. A cognitive overlay for AI. pip install linafish

In the telegraph era, you paid for every word. In the AI era, we do it again. Cablegram minimizes tokens while preserving meaning.

Rust compression core for Glasik Notation — rolling hash tokenizer, codon table, sliding window dictionary, PyO3 Python bindings

Semantic codec with rate-distortion optimization and renderer folding. #1 on comma.ai's video compression leaderboard (Oct 2026)

Codex skill for token-efficient Spanish technical answers with semantic compression, integrity guards, and reproducible benchmarks.

This repository explores the transition from probabilistic AI systems toward epistemic architectures capable of generating reproducible knowledge.

PhoLine — client-side PhO channel codec for LLM chats. Information filter, not IPA. By Christian Heinrich Hohlfeld, ORCID 0009-0003-6634-9045.

Capture clean, customizable code screenshots quickly with an open-source tool that requires no signup, backend, or setup.

UTF-8 language model compression achieving ~66% token reduction while preserving semantic meaning. Enables efficient AI on resource-constrained hardware.

Rust-first semantic compression toolkit: extract text from any file type into compact, LLM- and human-friendly representations.

Risk-aware communication firewall for multi-agent systems: target-relevant compression, observation refinement, uncertainty cascade detection, and verifier placement.