Show HN: Price Dashboard for LLM Inference
TL;DR
A developer built a lightweight price history tracker for LLM inference covering more than 100 platforms. It grew out of repeatedly hunting for the cheapest provider for models like DeepSeek and K3 after running out of quota on Claude Code and Codex. The tracker refreshes every six hours, so it records how prices move over time instead of showing only a snapshot. Updates are available via an Atom feed and the project is open to contributions.
Nauti's Take
Price transparency is a real advantage for teams, because inference costs finally become comparable over time instead of only at the moment of signup. The limit sits in the data itself: list prices say little about latency, rate limits or the actual quality a provider delivers.
The payoff shows up at serious volume, far less so on a fixed subscription.
Sources