We've spent the last few months building Infere, an AI routing and observability platform, and today we're putting it in front of real developers. We'd love for you to be the ones to stress-test it.
It started the way these things usually do, we were annoyed. Every AI feature we built meant another provider SDK, another auth quirk, another error format, another dashboard we'd have to open to find out what anything cost. Switching a model was a sprint. Nobody could answer "what did the support bot cost us this month" without a spreadsheet and a guess.
So we built Infere: one OpenAI-compatible endpoint we point our apps at, with a layer around it that picks the right model for each request, versions prompts like they were code, scores them with evaluators, and logs every single call so the cost is never a mystery again.
Point it at our endpoint, and routing, budgets, and observability kick in on your first request. Prepaid credits, pay as you go, credits never expire.
We're not going to pretend it's finished. We're still in beta. But it works, it's live, and we want people to actually use it against their real traffic, that's the feedback we can't invent.
Try it, break it, and tell us what's missing from the request path. We read everything.
Top comments (0)