token-reduction

[EMNLP 2024 & AAAI 2026] A powerful toolkit for compressing large models including LLMs, VLMs, and video generative models.

~97% token reduction for AI coding sessions — zero deps, 33 languages, MCP server

Save tokens. Maximize context, Safely

CLI proxy that reduces LLM token usage by 60-90%. Declarative YAML filters for Claude Code, Cursor, Copilot, Gemini. rtk alternative in Go.

The code for "TokenPacker: Efficient Visual Projector for Multimodal LLM", IJCV2025

A discovery and compression tool for your Python codebase. Creates a knowledge graph for a LLM context window, efficiently outlining your project | Code structure visualization | LLM Context Window Efficiency | Static analysis for AI | Large Language Model tooling #LLM #AI #Python #CodeAnalysis #ContextWindow #DeveloperTools

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.

A lightweight tool to optimize your Javascript / Typescript project for LLM context windows by using a knowledge graph | AI code understanding | LLM context enhancement | Code structure visualization | Static analysis for AI | Large Language Model tooling #LLM #AI #JavaScript #TypeScript #CodeAnalysis #ContextWindow #DeveloperTools

[CVPR 2025] PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models

ZON → 35-70% cheaper LLM prompts than JSON/TOON. Zero overhead.

⚡ Cut Claude token usage by 90%+ — free, open-source, local-first context compression for Claude Code. Hybrid RAG (BM25 + ONNX vectors), AST chunking, reranking. No API needed.

[AAAI 2026] Global Compression Commander: Plug-and-Play Inference Acceleration for High-Resolution Large Vision-Language Models

token-ninja routes deterministic shell commands locally — zero LLM calls, ~19µs latency. Works silently inside AI tools via MCP.

A deterministic, zero-dependency prompt-pruning layer for long-running LLM conversations — expires stale tool state, collapses duplicate context, and proves it never drops a fact a later turn still depends on.

A lightweight tool to optimize your C# project for LLM context windows by using a knowledge graph | Code structure visualization | Static analysis for AI | Large Language Model tooling | .NET ecosystem support #LLM #AI #CSharp #DotNet #CodeAnalysis #ContextWindow #DeveloperTools

MCP server for .NET: typecheck AI agent edits against the compiler, resolve DI and route wiring from Roslyn, scope PRs to the files that matter.

Persistent memory for AI coding agents. Your agent learns your codebase the way a senior engineer would — what files go together, what you usually edit next. Works with Claude Code, Cursor, Cline, Continue. 100% local. (Side effect: 12-50× cheaper code questions.)

A discovery and compression tool for your Java codebase. Creates a knowledge graph for a LLM context window, efficiently outlining your project #LLM #AI #Java #CodeAnalysis #ContextWindow #DeveloperTools #StaticAnalysis #CodeVisualization

DUET-VLM: Dual stage Unified Efficient Token reduction for VLM Training and Inference

Verdict-first output for AI coding agents. Tiny prompt + installer for Claude Code, Codex, Gemini, Cursor, opencode, and 30+ agents.