OpenAI Unveils Coding-Focused GPT-4.1 While Phasing Out GPT-4.5

OpenAI launched the GPT-4.1 model family, focusing on improved coding and instruction adherence. The family includes three variants: GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano, all accessible via API. GPT-4.1 shows strong performance in long video comprehension but lags behind competitors in software engineering benchmarks. OpenAI is retiring GPT-4.5, citing cost-effectiveness improvements with GPT-4.1. GPT-4.1 offers a tiered pricing structure based on model size and usage. The nano variant is highlighted as OpenAI's fastest and most affordable model. All GPT-4.1 models have a June 2024 knowledge cutoff for updated information. The models are not currently integrated into ChatGPT. OpenAI emphasizes that GPT-4.1 provides comparable or better performance at reduced costs compared to its predecessor.