{
  "slug": "helicone",
  "name": "Helicone",
  "url": "https://www.helicone.ai",
  "cat": "ai",
  "group": "ai",
  "tagline": "LLM observability proxy",
  "price": "Hobby free (10,000 requests, 1GB, 7-day retention), Pro $79/mo, Team $799/mo, Enterprise custom, all usage-based above included requests. Open source and self-hostable; 50% startup discount.",
  "tier": "mixed",
  "pick": false,
  "what": "Open-source LLM <b>observability</b> that works as a one-line proxy: change your base URL and every request is logged with cost, latency and prompt content. It adds caching, rate limiting and per-user spend tracking, with an async SDK if you would rather not proxy.",
  "why": "The lowest-effort way to find out where your token spend actually goes — no instrumentation, no framework buy-in, just a base URL swap. Per-user cost tracking is what makes usage-based billing possible.",
  "warn": "Proxy mode puts a third party in your critical path, adding latency and an outage dependency — use the async logger in production, or self-host. Free retention is only 7 days.",
  "install": "npm i @helicone/helpers",
  "category": {
    "key": "ai",
    "name": "Model APIs, routers & observability",
    "desc": "Providers, gateways, inference hosts and eval layers — constantly conflated, listed separately here.",
    "group": "ai",
    "groupName": "Put AI inside it"
  },
  "freshness": null,
  "health": {
    "ok": true,
    "status_code": 200,
    "final_url": "https://www.helicone.ai/",
    "checked_at": "2026-08-18T18:45:10.648+00:00"
  },
  "corrected_at": null,
  "upstream_changed_at": null
}