OpenAI wants developers using its new GPT-4.1 models - but how do they compare to Claude and Gemini on coding tasks?

OpenAI has launched the GPT-4.1 family of AI models aimed at developers, boasting improved coding skills and contextual processing abilities. This multi-modal model can handle prompts of up to a million tokens, addressing the challenge of detail extraction in extensive texts. The release includes two smaller variants: GPT-4.1 mini, which matches the performance of GPT-4o but with lower latency, and GPT-4.1 nano, optimized for fast text-based tasks. GPT-4.1 shows a notable 21.4% improvement in coding benchmarks over GPT-4o, emphasizing its potential in professional applications, such as code analysis and document processing.

OpenAI's GPT-4.1 offers significant advancements in coding capabilities, with a 21.4% improvement over its predecessor in handling real-world coding problems from GitHub data.

With support for up to a million tokens, GPT-4.1 is designed to excel in professional settings by effectively processing large documents, enhancing user interaction.

Read at IT Pro