ai-command-center
v0.2.1
Published
Plug-and-play LLM gateway + consolidated usage & cost dashboard for every AI project you run - any language, one dashboard, one command.
Maintainers
Readme
ai-command-center
One gateway, every AI project, one dashboard.
A zero-dependency LLM gateway with a self-hosted usage and cost dashboard. Point any project at it - any language, one command - and see tokens, cost, latency, and errors for every AI product in one place. Your API keys pass straight through, and prompt and response bodies are never stored.
npx ai-command-center # gateway + dashboard on http://localhost:4321
npx ai-command-center demo # seed 14 days of sample data to explore firstIntegrate by changing one base URL. Your key never moves:
from openai import OpenAI
client = OpenAI(base_url="http://localhost:4321/p/invoice-bot/openai/v1")That is the whole integration. It works with OpenAI, Anthropic, Gemini,
OpenRouter, Mistral, DeepSeek, xAI, Groq, Together, Ollama, and any
OpenAI-compatible endpoint - from any language, or with zero code change via
OPENAI_BASE_URL / ANTHROPIC_BASE_URL / GOOGLE_GEMINI_BASE_URL.
The dashboard has five views: cost and usage (in rupees, dollars, or euros at live rates, exact from real token counts), traces and sessions, prompt versions, model comparison, and budgets with anomaly alerts. Optional login, teams, and per-project gateway keys let a team see only their own projects. Everything is local append-only JSONL - no database, no runtime dependencies. It is not a tracing, eval, or routing platform; it is the cost dashboard that just runs.
Docs & live demo
https://aicommandcenter.vercel.app - quickstart, per-language integration, config, auth, the HTTP API, providers, self-hosting, security, and an honest comparison with Langfuse, Helicone, LangSmith, LiteLLM, and Portkey.
CLI
ai-command-center start | demo | clear | stats | snippets | user
--port --host --data-dir --config --preset --no-authSource, issues, contributing: https://github.com/adityasarade/ai-command-center · MIT.
