OpenAI has launches GPT-4o mini, a smaller and cheaper AI model that outperforms other small AI models. This new model is designed to be faster and more cost-efficient, making it an attractive option for developers and businesses. Here are some key features and benefits of GPT-4o mini:
- Faster and More Cost-Efficient: GPT-4o mini is faster and more cost-efficient than its predecessor, GPT-3.5 Turbo . It has a median output speed of 202 tokens per second, making it more than 2X faster than GPT-4o and GPT-3.5 Turbo.
- Better Performance: GPT-4o mini outperforms other small AI models on reasoning tasks involving text and vision. It scores 82% on MMLU, a benchmark to measure reasoning, compared to 79% for Gemini 1.5 Flash and 75% for Claude 3 Haiku.
- Cheaper: GPT-4o mini is significantly cheaper to run than its previous frontier models, and more than 60% cheaper than GPT-3.5 Turbo. It is priced at 15 cents per million input tokens and 60 cents per million output tokens .
- Supports Text and Vision: GPT-4o mini supports text and vision in the API, and OpenAI plans to add video and audio capabilities in the future.
- Replacing GPT-3.5 Turbo: GPT-4o mini will replace GPT-3.5 Turbo as the smallest model OpenAI offers. GPT-3.5 Turbo will still be available for developers via the API, but it will be retired at some point.
- Available Now: GPT-4o mini is available now for developers and will be available for ChatGPT users on Free, Plus, and Team plans starting today. Enterprise users will get access next week.
GPT-4o Mini: Built with Robust Safety Measures
OpenAI has prioritized safety in the development of GPT-4o mini, integrating rigorous safety protocols to ensure the model's reliability and trustworthiness. Key safety features include:
1. Pre-training filtering: Harmful content is meticulously filtered out during the pre-training process to prevent the model from learning inappropriate or toxic material.
2. Reinforcement Learning with Human Feedback (RLHF): This technique is employed to align the model's behavior with established safety policies, guaranteeing that GPT-4o mini adheres to ethical standards.
3. Expert testing and risk mitigation: Over 70 external experts have thoroughly tested the model to identify and address potential risks, ensuring that GPT-4o mini is a safe and reliable tool for users.
These comprehensive safety measures demonstrate OpenAI's commitment to responsible AI development, making GPT-4o mini a trustworthy and robust model for various applications.
Get Ready to Harness the Power of GPT-4o Mini
Exciting news! GPT-4o mini is now available through various APIs, including Assistants API, Chat Completions API, and Batch API. This means developers can start exploring its capabilities and integrating it into their applications.
Starting today, ChatGPT users on Free, Plus, and Team plans can access GPT-4o mini, with enterprise users getting access next week. But that's not all - OpenAI is also planning to introduce fine-tuning capabilities for GPT-4o mini soon, allowing for even more customization and flexibility.
OpenAI's goal is to make advanced AI more accessible and affordable for everyone. With GPT-4o mini, they're taking a significant step towards achieving this goal. By reducing costs and enhancing AI capabilities, OpenAI is paving the way for powerful AI to become an integral part of our everyday live
