sqz - Context Compression 作成者: Ojus Chugh

sqz compresses text you paste into AI chat tools, reducing token usage and saving costs.

When you paste 2000+ characters into any supported AI chat site, sqz shows a compression preview banner with before/after token counts. Click "Compress" to apply, or "Skip" to keep the original.

How it works:
- 16-pass compression engine
- Works on: ChatGPT, Claude.ai, Gemini, Grok, Perplexity

Privacy:
- All processing happens locally in your browser
- Zero external network requests
- No telemetry, no analytics, no data collection
- Open source: https://github.com/ojuschugh1/sqz

Part of the sqz context intelligence layer - also available as a CLI tool (cargo install sqz-cli) and MCP server.