OpenAI has expanded its advanced GPT-4.1 and GPT-4.1 mini AI models to ChatGPT, bringing powerful new coding capabilities directly to chatbot users. Previously limited to API access, these models are now available in the ChatGPT interface, offering enhanced performance in programming and instruction-following tasks.
The GPT-4.1 model is now accessible to ChatGPT Plus, Pro, and Team subscribers via the “more models” dropdown menu. Enterprise and Edu users will gain access in the coming weeks, while free-tier users will receive the GPT-4.1 mini model, which replaces the GPT-4o mini.
OpenAI claims GPT-4.1 surpasses GPT-4o and 4.5 in coding benchmarks, scoring 54.6% on the SWE-bench Verified benchmark. It also showed improved results in instruction-following and multimodal understanding, including a 72% score on the Video-MME benchmark. The model supports a context window of up to one million tokens.
The 4.1 mini model, while smaller, also outperforms GPT-4o in several intelligence benchmarks, offering 50% lower latency and 83% reduced costs—making it a more efficient choice for developers and users alike.