A lightweight registry of LLM model information, like name and context sizes, for building AI-powered apps.
Tree-shakeable static models.dev catalog split by provider for TokenLens.
Core types and registry utilities for TokenLens (model metadata).
Tokenizer utilities for Tokenlens.
Typed client for models.dev to fetch model catalogs with friendly errors.
Helpers for context windows, usage normalization, compaction, and cost estimation.
Bun-first local token analytics for coding agents.
Local AI coding token usage dashboard
See exactly what's eating your Claude Code context window — and stop it.
React hooks for TokenLens — real-time token counting in your components.
Local-only Claude Code spend analyzer. Reads your session transcripts and surfaces token waste. No signup, no cloud, no telemetry.
GEPA (Genetic-Pareto) optimizer for TypeScript - Multi-objective optimization through reflective evolution, with first-class support for AI SDK's generateObject
A lightweight Typescript library to track and visualize token usage across multiple AI model providers.