The Token Archive
Features

Cut input tokens. Keep the same models.

Drop-in OpenAI-compatible proxy. Compress before upstream. Meter every call.

Compression

Shrink prompts before they hit OpenAI / OpenRouter / Anthropic.

Live meter

Original vs compressed tokens and savings % on every request.

One-line install

Swap base URL: Cursor, Continue, Cline, Roo, Hermes, Aider, Open WebUI, SDKs, LangChain. Full matrix in the docs.

Modes + BYOK

Lossless / balanced / aggressive. Your provider keys stay yours.