OpenMeter

OpenMeter fills a specific niche: real-time usage metering for AI/API companies that need to bill on token counts or API calls. Its event-streaming architecture handles millions of events per second with low latency, which is critical for usage gating and quotas. The no-code product catalog lets product teams change pricing without touching code, a clear win over building in-house. The September 2025 acquisition by Kong adds enterprise backing but also noise. The cloud product rebranded to Kong Metering & Billing, while the open-source core stays standalone. If you're already a Kong user, integration may be smoother; if not, the brand shift could be a distraction. We'd reach for OpenMeter when you need to meter per-millisecond compute (like Trigger.dev does) or per-LLM-call billing. It's less suited for flat-rate subscriptions or legacy ERP billing. The self-hosted option is powerful but requires Kafka or Postgres for event buffering, adding ops overhead. Compared to Metronome, OpenMeter is more developer-friendly with SDKs and collectors. Stripe's usage billing is simpler but lacks real-time enforcement. OpenMeter gives you both flexibility and openness, at the cost of some polish in the cloud UI. Where it bites: the cloud pricing page is now just a redirect to Kong, with no explicit tiers. You have to contact sales for Pro. For self-hosted, it's free but you own the ops. Documentation is decent but could be more comprehensive for advanced use cases like multi-currency or tax rules beyond Stripe Tax. Overall, if you need real-time metering for a usage-based product and have developer resources, OpenMeter is a solid pick. Just account for the Kong transition and plan your cloud vs. self-hosted path early.