OpenAI Launches Cheaper Small AI Model GPT-4o Mini

GPT 4-o Mini

OpenAI, the creator of ChatGPT, announced on today the launch of GPT-4o mini, a smaller and more cost-efficient AI model. This new development aims to make the company’s technology more affordable and less energy-intensive, allowing OpenAI to target a broader range of customers.

As the market leader in AI software, Microsoft-backed OpenAI has been focused on making it cheaper and faster for developers to build applications based on its model. This strategic move comes at a time when deep-pocketed rivals, such as Meta and Google, are rushing to grab a bigger share of the market.

Starting today, ChatGPT’s Free, Plus, and Team users will gain access to GPT-4o mini, which has knowledge up to October 2023. This new model will replace GPT-3.5 Turbo. Meanwhile, enterprise users will be able to access GPT-4o mini starting next week, according to OpenAI.

OpenAI has introduced a groundbreaking development in the field of artificial intelligence: GPT-4o mini. This new model is touted as the most cost-efficient small model available, designed to make AI more accessible and affordable for a wider range of applications. Here’s a deep dive into what makes GPT-4o mini a game-changer in the AI landscape.

Unmatched Cost Efficiency

One of the most striking features of GPT-4o mini is its affordability. Priced at just 15 cents per million input tokens and 60 cents per million output tokens, it is significantly cheaper than previous models. This dramatic reduction in cost—more than 60% cheaper than GPT-3.5 Turbo—opens up new possibilities for developers and businesses looking to integrate AI into their operations without breaking the bank.

See also  Using SPDTransNet Transformers for Better Sleep Analysis

Performance and Versatility

Despite its small size, GPT-4o mini does not compromise on performance. It outperforms its predecessors and other small models across various benchmarks. For instance, it scores 82% on the MMLU, a benchmark for textual intelligence and reasoning, which is higher than both Gemini Flash (77.9%) and Claude Haiku (73.8%).

GPT-4o Mini’s Broad Application Scope

Thanks to its low cost and high efficiency, GPT-4o mini is ideal for a broad range of tasks. It can handle applications that require chaining or parallelizing multiple model calls, managing large volumes of context (such as full code bases or conversation histories), and providing fast, real-time text responses for customer support chatbots.

GPT-4o mini supports text and vision in the API, with plans to include text, image, video, and audio inputs and outputs in the future. With a context window of 128K tokens and support for up to 16K output tokens per request, it stands ready to tackle complex and extensive tasks.

GPT-4o Mini Superior Multimodal Reasoning

GPT-4o mini is not just about textual intelligence; it also excels in multimodal reasoning. It has been evaluated on several key benchmarks, proving its superiority in both mathematical reasoning and coding tasks. For example, it scored 87.0% on MGSM (measuring math reasoning) and 87.2% on HumanEval (measuring coding performance), outperforming other models like Gemini Flash and Claude Haiku.

Built-in Safety and Reliability

Safety and reliability are at the core of GPT-4o mini’s design. OpenAI has incorporated extensive safety mitigations, informed by both automated and human evaluations. More than 70 external experts in fields such as social psychology and misinformation have tested the model to identify potential risks. These insights have been instrumental in enhancing the model’s safety, making it more resistant to jailbreaks, prompt injections, and system prompt extractions.

See also  OpenAI Launches DALL·E 3: A Revolution in Text-to-Image Generation

Democratizing AI Access

Starting today, GPT-4o mini will be available to Free, Plus, and Team users in ChatGPT, replacing GPT-3.5. Enterprise users will gain access next week, further aligning with OpenAI’s mission to democratize AI access. This move is set to pave the way for developers to build and scale powerful AI applications more efficiently and affordably.

The Future of AI

The introduction of GPT-4o mini marks a significant milestone in AI development. It represents OpenAI’s ongoing commitment to reducing costs while enhancing model capabilities, ensuring AI becomes seamlessly integrated into our daily digital experiences. As OpenAI continues to lead the way, the future of AI looks more accessible, reliable and embedded in our everyday lives.

For more details, visit the official announcement here.

Get ready to dive into a world of AI news, reviews, and tips at Wicked Sciences! If you’ve been searching the internet for the latest insights on artificial intelligence, look no further. We understand that staying up to date with the ever-evolving field of AI can be a challenge, but Wicked Science is here to make it easier. Our website is packed with captivating articles and informative content that will keep you informed about the latest trends, breakthroughs, and applications in the world of AI. Whether you’re a seasoned AI enthusiast or just starting your journey, Wicked Science is your go-to destination for all things AI. Discover more by visiting our website today and unlock a world of fascinating AI knowledge.

About Author

Teacher, programmer, AI advocate, fan of One Piece and pretends to know how to cook. Michael graduated Computer Science and in the years 2019 and 2020 he was involved in several projects coordinated by the municipal education department, where the focus was to introduce students from the public network to the world of programming and robotics. Today he is a writer at Wicked Sciences, but says that his heart will always belong to Python.