OpenAI has expanded its GPT-4.1 series of artificial intelligence (AI) models to include ChatGPT, previously available only through the company’s application programming interface (API). On Thursday, the San Francisco-based AI developer announced that the new models will now be integrated into the chatbot, specifically the GPT-4.1 and GPT-4.1 mini models. OpenAI highlighted that these updated models demonstrate improved capabilities in coding-related tasks and serve as an alternative to the o3 model.
OpenAI’s GPT-4.1 AI Models Are Now Coming to ChatGPT
In a statement on X (formerly Twitter), OpenAI’s official account shared that the GPT-4.1 AI model will now be accessible within the ChatGPT platform. The company emphasized the model’s strengths in coding tasks and its proficiency in following instructions. ChatGPT Plus, Pro, and Team users will find this option in the “more models” dropdown menu.
Enterprise and Edu subscribers can expect access to this model in the upcoming weeks, while users on the free tier will soon be able to utilize the GPT-4.1 mini model, which will replace the existing GPT-4o mini model.
Upon its launch, OpenAI reported that the GPT-4.1 series outperformed both the GPT-4o and GPT-4.5 models in coding capabilities. According to internal evaluations, GPT-4.1 achieved a score of 54.6 percent on the SWE-bench Verified benchmark.
In addition, it reportedly scored 38.3 percent on the MultiChallenge benchmark for instruction following and an impressive 72 percent on the Video-MME benchmark, which assesses multimodal long context understanding. The models in the GPT-4.1 series can handle context windows of up to one million tokens.
While the GPT-4.1 remains the flagship model, the GPT-4.1 mini also reportedly exceeds GPT-4o in various benchmarks, including intelligence assessments. OpenAI noted that this mini model reduces latency by 50 percent and is 83 percent more cost-effective compared to its predecessor.
