ChatGPT Model Selector
The ChatGPT model selector allows users to choose from various models based on their subscription plan. Models include GPT-4.5, o3-mini, o1, and GPT-4o, each with specific capabilities and limitations.

The ChatGPT model selector allows users to switch between different models based on their subscription plan and specific needs. Users with paid plans such as Plus, Pro, Team, or Enterprise can choose from various models, including GPT-4o, o1, o3-mini, and o3-mini-high. GPT-4o-mini and GPT-4 are listed under 'Other models'. Plus, Pro, or Team subscribers also have access to GPT-4.5, which will soon be available for Enterprise and Edu. Pro plan users can view the o1 pro mode.
GPT-4.5 is currently the most advanced model in preview. It has improved both pre-training and post-training phases, enhancing the ability to recognize patterns and generate creative ideas without explicit reasoning. Initial tests indicate that interacting with GPT-4.5 feels more natural. With its extensive knowledge base and ability to follow user intent, it is suitable for improving texts, programming tasks, and solving practical problems, with a reduced tendency to 'hallucinate'.
The o3-mini model is designed for advanced reasoning, excelling in programming and complex tasks. It combines speed, efficiency, and flexibility, making it ideal for both development and production. It shows programming and reasoning capabilities comparable to or sometimes exceeding those of the o1 model, with faster responses and lower costs. It is recommended as the default model for small-scale reasoning and supports web searches to provide updated answers, but does not handle visual tasks.
The o1 and o1-mini models are specifically designed to solve complex problems in areas such as research, strategy, programming, mathematics, and science. While GPT-4o remains the most versatile model, the o1 series models are particularly suited for advanced tasks. However, they lack some advanced features such as memory, custom instructions, and file uploads unless GPT-4o is used.
GPT-4o is OpenAI's flagship model, capable of reasoning simultaneously on audio, images, and text in real-time. It is available in ChatGPT and via API, supporting advanced functions such as data analysis and file uploads. GPT-4o is accessible to Free, Plus, Pro, Team, and Enterprise users. The GPT-4o mini model is the fastest and suitable for everyday tasks, available to all users and automatically used by Free users who reach the limits of GPT-4o. It does not support data analysis and file uploads.
GPT-4 is ideal for creative tasks and advanced reasoning. It is available for Plus, Team, and Enterprise users and also allows image generation with DALLE 3. Usage limits and API data vary based on the subscribed plan, with a higher maximum for Pro and Enterprise users. It is important to note that data sent via API is not used for model training without explicit user consent. ChatGPT offers different levels of access, with specific advantages and usage limits tailored to each plan.