
An agent gateway that routes coding-agent requests under token budgets, compresses tokens and tracks team usage.
An agent gateway that routes coding-agent requests under token budgets, compresses tokens and tracks team usage.
Edgee is an Agent Gateway that sits between coding agents and LLM providers and applies routing, compression and observability to every request. Teams set token budgets per developer, squad or org, and as a budget fills Edgee reroutes requests to open weight models mid-task. Token compression trims tool results at the gateway to reduce billed tokens, and dashboards attribute cost to people, repos and PRs. It works with Claude Code, Codex, Copilot, OpenCode and Cursor, supports bring-your-own keys, and can be deployed on-premise.
An agent gateway that routes coding-agent requests under token budgets, compresses tokens and tracks team usage.
Edgee works by combining Budget-driven Routing: Set token budgets per developer, squad or org and reroute to cheaper models mid-task as budgets fill., Token Compression: Trim tool results on input and enforce brevity on output so fewer tokens are billed for the same context., Team Observability: Attribute cost and usage per developer, repo, PR and model., Open Weight Models: GLM, Kimi and MiniMax served for a flat monthly price as routing fallbacks., Bring Your Own Keys: Use Edgee's keys or your own provider keys with 0% BYOK commission. to help users with Coding Agent Cost Control: Cap spend for Claude Code or Codex per developer or squad., Provider Failover: Continue on the next model when a provider fails or a plan cap is hit., Spend Attribution: Show engineering leaders cost per PR and repo., Private Deployments: Keep prompts and provider keys inside your network., Claude Plan Extension: Use other models inside Claude Code while keeping your Claude plan..
Key features include Budget-driven Routing: Set token budgets per developer, squad or org and reroute to cheaper models mid-task as budgets fill., Token Compression: Trim tool results on input and enforce brevity on output so fewer tokens are billed for the same context., Team Observability: Attribute cost and usage per developer, repo, PR and model., Open Weight Models: GLM, Kimi and MiniMax served for a flat monthly price as routing fallbacks., Bring Your Own Keys: Use Edgee's keys or your own provider keys with 0% BYOK commission..
Edgee is useful for anyone interested in Coding Agent Cost Control: Cap spend for Claude Code or Codex per developer or squad., Provider Failover: Continue on the next model when a provider fails or a plan cap is hit., Spend Attribution: Show engineering leaders cost per PR and repo., Private Deployments: Keep prompts and provider keys inside your network., Claude Plan Extension: Use other models inside Claude Code while keeping your Claude plan..
Edgee offers a free tier with paid plans for advanced features.
Visit https://www.edgee.ai to sign up and explore Edgee.
Compare Edgee: vs BAND for Developers · vs NitroStack · vs GBrain · vs MentionBird