OpenAI has introduced GPT-4o mini, a compact yet potent AI model aimed at democratizing access to artificial intelligence with its affordability and efficiency. Priced at just 15 cents per million input tokens and 60 cents per million output tokens, GPT-4o mini offers a cost-effective solution compared to previous models like GPT-3.5 Turbo.
According to OpenAI's blog, GPT-4o mini delivers impressive performance metrics, achieving an 82 percent score on the MMLU benchmark and surpassing GPT-41 on the LMSYS leaderboard for chat preferences. This model is tailored for applications requiring multiple model calls, extensive context handling, or real-time text interactions, such as customer support chatbots.
Currently, GPT-4o mini supports text and vision inputs, with future plans to incorporate capabilities for image, video, and audio inputs and outputs. It boasts a context window of 128K tokens and supports up to 16K output tokens per request, leveraging knowledge updates until October 2023. Its enhanced tokenizer enhances cost-effectiveness, especially for non-English text handling.
The model excels in both academic and practical domains, demonstrating superior performance in reasoning, mathematical tasks, and coding challenges. For instance, GPT-4o mini achieved 87 percent in mathematical reasoning and 87.2 percent in coding performance on benchmarks such as MGSM and HumanEval, respectively.
To ensure safety and reliability, OpenAI has implemented robust measures in GPT-4o mini, including pre-training content filtering for harmful material and reinforcement learning with human feedback (RLHF) to align the model's behavior with safety protocols. Over 70 external experts have rigorously tested the model to identify and mitigate potential risks.
GPT-4o mini is accessible through OpenAI's Assistants API, Chat Completions API, and Batch API. It is initially available to Free, Plus, and Team users on ChatGPT, with enterprise users gaining access shortly. OpenAI plans to introduce fine-tuning capabilities for GPT-4o mini soon.
With GPT-4o mini, OpenAI aims to lower costs while enhancing AI capabilities, paving the way for integrating advanced intelligence into everyday applications. This initiative makes powerful AI more accessible and affordable for developers and users, marking a significant stride towards widespread adoption and innovation in artificial intelligence.
