robinbanner

3 ms·

1. 1.▲
   by robinbanner 8mo ago

   Backstory: I was building a customer support AI for a client last year. We started with Claude Opus for everything because it worked great. The bill was $250/month for maybe 10K conversations. Then I looked at the actual queries. 70% w
2. 2.▲
   Show HN: API router that picks the cheapest model that fits each query(komilion.com)

   1 points by robinbanner 8mo ago|1 comments
3. 3.▲
   by robinbanner 9mo ago

   Nice dashboard. Cost visibility is the first step — you cannot optimize what you cannot measure. We have found that the natural next step after tracking is automated routing. Once you see that 70% of your frontier model calls could have bee
4. 4.▲
   by robinbanner 9mo ago

   This is a nice approach — using model internals to predict success before committing to full generation. We took a different path at Komilion: classify the prompt upfront (regex fast path + lightweight LLM classifier) and route to the cheap
5. 5.▲
   by robinbanner 9mo ago

   Interesting approach. I built something similar (Komilion) — a model router that classifies each request and routes to the cheapest model that handles it well. A few things I learned building this: 1. Price-only routing is not enough. You n