GPT-4o mini
A cost-effective intelligent model
- Supports text and visual input and output, and will be expanded to include images, videos, and audio in the future.
- A context window with 128K tokens, supporting up to 16K output tokens per request.
- The knowledge deadline is October 2023, and it is capable of processing non English texts.
- Has performed well in academic benchmark tests, particularly in text intelligence and multimodal reasoning.
- Built in security measures ensure that model behavior complies with policies through pre training and post training filtering and alignment techniques.
- Support chaining or parallelizing multiple model calls, suitable for applications that require a large amount of context or fast response.
- The price is 15 cents per million input tokens and 60 cents per million output tokens, significantly lower than previous frontier models.
Product Details
GPT-4o mini is a highly cost-effective small intelligent model launched by OpenAI. It surpasses other small models in multimodal reasoning and text intelligence, and supports the same language range as GPT-4o. This model performs well in mathematical reasoning and coding tasks, capable of processing large amounts of contextual information and supporting fast, real-time text response. The launch of GPT-4o mini aims to make intelligent technology more widely applied in various application scenarios, reduce costs, and improve accessibility.
