Helicone AI

Helicone is an observability and routing platform for teams building applications on large language models. It sits between an application and the model providers it calls, logging every request so developers can trace sessions, debug failures, measure cost and latency, and analyze how real traffic exercises the system. The same layer doubles as an AI gateway, adding caching, rate limits and automatic fallbacks between providers so an app keeps working when one endpoint degrades. It is aimed at engineering teams who integrate with a single line of code and then want the full picture: tracing, user analytics, custom properties, a query language over logs, alerts and scheduled reports. A prompt and testing suite covers a playground, prompts, scores, datasets and webhooks, so experiments and evaluations live alongside production traffic. The platform is open source, and the plans differ mainly in retention, ingestion rate and API throughput rather than in which features are available at all. Pricing has four tiers. Hobby is free and includes 10,000 requests per month, 1 GB of storage and one seat in one organization. Pro is $79 per month and adds unlimited seats, alerts and reports, and the query language. Team is $799 per month for five organizations, SOC 2 and HIPAA compliance, and a private Slack channel. Enterprise is quoted on request and covers SAML single sign on, on premises deployment and a custom agreement. Usage based charges for requests and storage apply above the included amounts on the paid tiers, each of which carries a seven day free trial. Separate discounts are published for startups, non profits, open source companies and students. The site carries a banner stating that Helicone has joined Mintlify.