Auriko - Trading desk for LLM calls

Nice launch! LLM cost optimization is exactly where a lot of teams need help right now.

Love the "trading desk for inference" framing—routing on cache behavior and real-time provider signals instead of just headline prices is exactly the kind of optimization most teams skip, and the zero-markup model makes it a no-brainer to try. Congrats on the launch! 🚀

The trading desk analogy is compelling but trading desks also have slippage, the cost of execution diverging from the expected price. What's the equivalent for LLM routing, like how often does Auriko route to a provider that then has a latency spike or quality degradation that negates the cost saving, and is there a real-time feedback loop that reroutes mid-session or only adjusts for future requests?

@zxy_action1 Congratulations and happy product hunt.

llm calls as a trading desk is such a clever framing 📈 30% savings is a real hook, congrats on #1

Optimizing for the expected cost of the full session instead of the cheapest individual request is the interesting part here. Does the routing model also account for context continuity beyond cache economics—for example, provider-specific differences that could cause subtle behavioral drift during a long agent run?

As more routing platforms start optimizing across the same inference providers, do you think the opportunity for cost arbitrage naturally shrinks over time—similar to how financial markets become more efficient—or do you expect new pricing inefficiencies to keep emerging as providers compete?