OpenAI introduces new o1 capability in OpenAI API

AI Impact Summary

OpenAI is introducing a new capability called o1 in the OpenAI API. This adds an additional model/feature option that could alter performance, latency, and cost profiles for prompts, embeddings, or other tasks depending on how o1 is implemented. Engineers should review the official docs for o1 endpoints, run a constrained pilot comparing it to current models, and adjust routing, monitoring, and pricing plans accordingly.