OpenAI Introduces GPT-4.1 Models Focused on Coding

OpenAI has launched a new family of models called GPT-4.1, which includes GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano. These models are designed to excel in coding and instruction following, featuring a 1-million-token context window that allows them to process approximately 750,000 words at once. The models are available through OpenAI's API but not on ChatGPT.

The GPT-4.1 models are optimized for real-world software engineering tasks, focusing on areas such as frontend coding, reliable format adherence, and consistent tool usage. OpenAI claims that the full GPT-4.1 model outperforms its predecessors on coding benchmarks like SWE-bench, although it slightly trails behind competitors like Google's Gemini 2.5 Pro and Anthropic's Claude 3.7 Sonnet.

Pricing for the models varies, with GPT-4.1 costing $2 per million input tokens and $8 per million output tokens. The mini and nano versions are more cost-effective, with GPT-4.1 nano being the fastest and cheapest model. Despite its capabilities, OpenAI acknowledges that GPT-4.1's reliability decreases with larger input sizes and that it may require more explicit prompts compared to previous models.