OpenAI starts rolling out GPT 4.1 model to all users: What is new

OpenAI has just made its latest GPT-4.1 and GPT-4.1 Mini models available in ChatGPT for everyone, the company announced on social media on Wednesday. If you’re subscribed to ChatGPT Plus, Pro, or Team, you’ll now see GPT-4.1 as an option. And the best part? GPT-4.1 Mini is free for all users, replacing the older GPT-4.0 Mini.

So, what’s the big deal? According to OpenAI spokesperson Shaokyi Amdo, GPT-4.1 shines when it comes to coding. Whether you’re writing new code or debugging existing scripts, the model follows instructions more accurately than its predecessor, GPT-4o, while also running faster than earlier reasoning-focused versions. That speed boost should help developers get their work done without those long waits.

Back in April, OpenAI quietly launched GPT-4.1 and its Mini sibling via its developer API. At that time, some AI researchers raised eyebrows because OpenAI didn’t include a full safety report, arguing the model isn’t a “frontier” or cutting-edge system that would require deep scrutiny. Johannes Heidecke, OpenAI’s Head of Safety Systems, reiterated this stance Wednesday, saying, “GPT-4.1 doesn’t introduce new ways of interacting or new modalities, and it doesn’t surpass our o3 models in intelligence. Its safety considerations differ from those of frontier models.”

OpenAI also announced it will publish its internal AI safety evaluations more regularly. These reports will live in the newly launched Safety Evaluations Hub, giving users and researchers a clearer window into how OpenAI tests and assesses model risks.