A token compression and AI gateway tool
Edgee is an AI gateway that compresses prompts at the edge to cut LLM input tokens and costs while preserving meaning. It routes requests across providers via an OpenAI-compatible API and adds tagging, alerts, and observability for governance.
Intelligent prompt token compression (up to 50% reduction)
Semantic preservation to keep intent and context intact
OpenAI-compatible gateway API for easy integration
Universal compatibility with OpenAI, Anthropic, Gemini, xAI, Mistral, and more
- Install via: curl https://test.edgee.ai | bash
- Create an API key and call the OpenAI-compatible send method with your model and input
- Add tags to requests to track spend by team, feature, or project
- Configure cost alerts to catch spikes before they impact budget
Reduces LLM spend and latency without changing application code
Works across multiple LLM providers under one normalized API
Strong cost tracking with tags, alerts, and end-to-end traffic visibility
Supports edge tooling and private/serverless model options