OpenAI, on Wednesday, unveiled an improved version of GPT-4 Turbo, which is now available via API (Application Programming Interface) and for ChatGPT Plus subscribers. GPT-4 Turbo is the companyโs latest generative AI model, which now has training data up to December 2023. The model is now capable of processing vision requests via JSON mode.
GPT-4 Turbo with Vision is now generally available in the API. Vision requests can now also use JSON mode and function calling.
Below are some great ways developers are building with vision. Drop yours in a reply ๐งต
โ OpenAI Developers (@OpenAIDevs) April 9, 2024
It still has a context window with up to 128,000 tokens, and using the new capability (via API), developers can deploy GPT-4 Turbo, which can analyze images, including food recognition from around the world. It is also more capable of generating computer code. Developers who have access to GPT-4 Turbo API can now implement the latest model within their apps and services.
The improved GPT-4 Turbo can now generate fully functional websites by producing real code. Additionally, it can read an image, understand the content and context, and produce meaningful output.
GPT-4 Turbo with vision capability via JSON mode is now available for ChatGPT Plus subscribers, which costs $20 a month. The same is also accessible via popular versions of ChatGPT on the ChatGPT store. Users can also update ChatGPT-4 Turbo via Microsoftโs Copilot by selecting creative or precise mode.
OpenAI also made it easy to access ChatGPT (free version) without logging into the service. Anyone with an active internet connection can access ChatGPT via the official website without signing up for the service. However, the free version of ChatGPT makes use of the GPT-3.5 model with limited knowledge compared to the latest GPT-4 Turbo.