Helicone
Open-source LLM observability, now in maintenance mode
Overview
Helicone is an AI gateway and LLM observability layer: change one base URL to log requests, costs and latency, with caching, rate limits and fallbacks. Mintlify acquired it in March 2026; it now runs in maintenance mode (security, bug and new-model fixes only) and new hosted sign-ups are disabled. The Apache-2.0 code remains self-hostable.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
What you can produce with Helicone
- Request logging for existing OpenAI or Anthropic integrations by changing the base URL
- LLM spend tracked by user, feature or environment with custom properties
- Caching of repeated prompts to cut cost and latency
- Rate limits and automatic fallbacks at the gateway
- Self-hosting of the Apache-2.0 stack with Docker
- Error-rate and latency monitoring across providers
- Migration support for existing customers from Mintlify
Helicone was one of the lowest-friction ways to see what your LLM calls cost. Since Mintlify acquired it in March 2026 it has been in maintenance mode, with security fixes and new-model support but no new features, and new hosted sign-ups have been switched off.
In practice it is now a tool for existing customers planning a migration, or for self-hosters running the Apache-2.0 stack who will maintain it themselves. For a new production deployment, choose an actively developed alternative.
What we found
Before evaluating Helicone, know that it is no longer a product you can start on. Mintlify announced on 3 March 2026 that it was acquiring Helicone, Inc., and the founders, Justin Torre and Cole Gottdank, joined Mintlify in San Francisco. Both companies said Helicone would run in maintenance mode, with security updates, bug fixes and new-model support continuing, and Mintlify said it would help every customer move to another platform. The same day, the team merged a change that switched off new sign-ups, and today both the US and EU sign-up pages say sign-ups are disabled. The pricing page still shows a free Hobby plan, Pro at $79 a month and Team at $799 a month, but new users cannot sign up for any of them.
The product itself is unchanged. It is an AI gateway and observability layer: change the base URL of an OpenAI- or Anthropic-style call, and every request is logged with its cost, latency and any custom properties you attach. Caching, rate limits, automatic fallbacks, sessions, prompt versioning and a SQL-like query language are included. At acquisition, Helicone said it had processed 14.2 trillion tokens for 16,000 organisations, its own figures.
The code is the part with a future. The GitHub repository is Apache-2.0, has not been archived and has Docker self-hosting docs. Commits since March have been maintenance and security fixes, including a September 2026 patch for a cross-tenant query bypass, with no new features. In May 2026 the hosted service lost about four days of log ingestion after AWS locked its account; Helicone says the proxy kept serving traffic and no data was lost.
Earlier versions of this page still described a generous free tier and an easy add to any stack. Existing customers should now plan a migration. Self-hosters who want provider-agnostic cost tracking can still run the open-source stack, but should expect to patch it themselves.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Helicone's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Helicone unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Helicone directly →
Spotted something out of date? Suggest an update →
Related reading
More in Code & Dev Tools
