OpneAI has just launched GPT-4o Mini, its most cost-efficient small AI Model. This model promises to broaden the scope of AI applications with its affordable pricing and powerful capabilities for the price.
GPT-4o mini is significantly more affordable than previous models. The GPT-4o mini is priced at 15 cents per million input tokens and 60 cents per million output tokens. This makes it an order of magnitude cheaper than its predecessors, including GPT-3.5 Turbo.
GPT-4o mini has already outperformed other small models on various benchmarks:
- Reasoning Tasks: Scores 82% on MMLU, surpassing Gemini Flash and Claude Haiku in reasoning tasks involving text and vision.
- Math and Coding Proficiency: Scores 87% on MGSM and 87.2% on HumanEval, leading its competitors in mathematical reasoning and coding tasks.
- Multimodal Reasoning: Achieved 59.4% on MMMU, outperforming Gemini Flash and Claude Haiku at multimodal reasoning evaluation.
The model has a context window of 128K tokens and supports text and vision inputs and outputs. GPT-4o mini is highly versatile. Future updates will include image, video, and audio inputs and output support.
GPT-4o mini excels in applications that require:
- Chaining or parallelizing multiple model calls.
- Handling large volumes of context.
- Providing fast, real-time text responses.
OpenAI has integrated strong safety measures into GPT-4o mini. The company has filtered out harmful content and applied reinforcement learning with human feedback (RLHF). The model also uses an instruction hierarchy method to resist jailbreaks and prompt injections, making it safer for large-scale applications.
GPT-4o mini is available to developers through various APIs. Free, Plus, and Team users in ChatGPT can access it immediately, with Enterprise users gaining access next week.
OpenAI seems committed to reducing costs while improving model capabilities. Fine-tuning for GPT-4o mini will be available soon, further expanding its usability.
GPT-4o mini has set a new standard in affordable and high-performing AI models. This mini-model allows for different applications and makes AI more accessible to developers and businesses. The future of AI looks more promising with this new model by OpenAI.
Nishant, the Product Growth Manager at Marktechpost, is interested in learning about artificial intelligence (AI), what it can do, and its development. His passion for trying something new and giving it a creative twist helps him intersect marketing with tech. He is assisting the company in leading toward growth and market recognition.