# Tokenledge

> Tokenledge is a drop-in token-compression proxy for AI coding agents (Claude Code, Cursor, Cline). It compresses tool output, logs, and file reads before they reach the model, cutting token usage 60-90% while preserving answer quality. No code changes required. Compression is reversible — original responses stay retrievable on demand.

## Product
- Solo: $10/mo, single developer, unlimited repos
- Power user: $20/mo, higher throughput ceiling, priority tuning
- Team: planned for later in 2026 — shared config, per-project spend visibility

## Status
Early access / waitlist stage as of July 2026. Not yet a registered company; pricing shown is a plan, not a live bill.

## Contact
Early access requests: see the waitlist link on https://tokenledge.vercel.app/