OpenAI just reduced its embedding API price significantly. It's a go-to for many developers, but it's not always the right choice. Slow response times can really erode user experience. In this analysis, I compared latency across popular APIs and models.