[OpenAI: GPT-5.2 Series Model Speed Increased by 40%, Inference Stack Optimized for API Users] According to Foresight News and official OpenAI developer updates, the speed of GPT-5.2 and GPT-5.2-Codex has now been increased by 40%. OpenAI has optimized the inference stack for all API users. OpenAI stated that this optimization uses the same model and weights, achieving lower latency.