OpenAI announced that the new GPT-4.1 model surpasses both GPT-4o and GPT-4o mini in various benchmarks, particularly excelling in agent-based coding tasks and long-context comprehension.
OpenAI has launched GPT-4.1, its newest flagship AI model, designed to excel in coding and instruction. Alongside it, the company introduced two smaller variants—GPT-4.1 mini and GPT-4.1 nano—available via API but not through ChatGPT.
According to OpenAI, the new models outperform GPT-4o and GPT-4o mini in multiple benchmarks, particularly in agent-based coding tasks and long-context understanding.
“GPT4.1 significantly improves upon GPT4o in coding tasks, offering better agent-driven solutions, frontend coding, fewer unnecessary edits, and more reliable formatting,” the company stated in a blog post.
Pricing for the new model is set at $2 per million input tokens and $8 per million output tokens. The mini and nano versions are designed for speed and affordability, with GPT-4.1 nano priced at just $0.10 per million input tokens and $0.40 per million output tokens—OpenAI’s most cost-effective model to date.
At the TED 2025 event, CEO Sam Altman shared that OpenAI’s tools have amassed 800 million users worldwide. COO Brad Lightcap highlighted India as the fastest-growing market for ChatGPT, especially since the launch of its Ghibli-inspired image-generation feature.
Within a week, over 130 million users had generated more than 700 million images. Altman expressed admiration for India’s swift AI adoption, stating, “What’s happening with AI in India right now is incredible.
The creativity explosion is outpacing the world.” He also noted that while it took ChatGPT five days to reach one million users after its public debut, last week, the platform added one million users every hour.
