Tracium is a developer-first observability layer for AI systems. It lets you track costs, trace requests, debug failures, compare prompts and models, and detect drift in real time, all across models, tools, and agents, in one place. Useful AI observability without complex setup, availabile in minutes and built for developers.
Hey PH 👋
Hunted this one because I've been using it on some of our products like Aqtos and is solving a big painpoint for us, especially on multi tenant cost and error handling analysis.
With just one line of code - tracium.trace() - Tracium.ai gets you full observability across your LLM calls, agents, and tool chains. Costs, errors, latency, drift. No config hell or complex dashboards.
If you're running anything AI in production and flying blind on what it's actually costing or where it's failing, Tracium.ai fixes that. There is a Free tier, no card required. And the maker @antonij_simonovski is here, ask him whatever interests you...
Hey Product Hunt! Antonij here, the founder of Tracium.
I built this after AI agents I was working on racked up huge bills with zero visibility into why. Every observability tool I tried wanted me to change my API client, add decorators to my functions. By the time I was done setting up monitoring I had lost time and still wasn't sure it was working.
So I built Tracium around one constraint: it has to work with one line of code or it's too hard.
You immediately see costs, errors, traces grouped by agent, and per-user breakdowns. Pricing tables update daily so you never have to maintain them yourself.
I'm talking to every user personally. If you try it and something feels off, reply here or reach out directly. I read everything.
Damjan Dano