llm-inference / app
CLI Proxy API Management Center
Capability: CLI Proxy API Management Center
Use it when
- You run a CLI Proxy API instance routing to Claude, Gemini, Codex, and OpenAI-compatible providers and need a UI to manage provider keys, model aliases, and quotas instead of hand-editing config.yaml
- Your self hosted LLM proxy needs OAuth or device-flow credentials for providers like Codex, Claude, Kimi, or Grok and you want to run those flows and upload auth JSON files from a browser
What it solves
Not the fit when
- Acting as the LLM proxy itself or forwarding any traffic
- Managing proxies other than CLIProxyAPI
- Programmatic automation; it is a browser UI over the Management API
- It is the web UI only; it is not a proxy and does not forward traffic - requires a running CLIProxyAPI backend >= 7.1.0
- Remote (non-localhost) access requires allow-remote-management to be enabled server side
- Management key is stored in browser localStorage with only lightweight obfuscation; treat as sensitive
- Older backends show some pages as unsupported (per-credential model lists, excluded models, logs)
Install
Bundled with CLI Proxy API >= 6.0.19: open http://<host>:<api_port>/management.html; or run standalone with bun install --frozen-lockfile && bun run dev
Invoke
Open management.html on the proxy's API port, enter the management key (sent as Authorization: Bearer), then use the Config, AI Providers, Auth Files, OAuth, Quota, and Logs pages
Alternatives
No reviewed alternatives recorded yet.