OpenAI Unveils Faster GPT-4 Turbo, Custom Chatbots, and New Assistants API
OpenAI launches GPT-4 Turbo with a 128k context window and lower costs, alongside new tools that let users build custom GPTs and developer assistants.
OpenAI unveils GPT-4 Turbo, a significantly upgraded AI model that features a 128k context window capable of processing over 300 pages of text at once. This new model updates its knowledge base to include world events up to April 2023 and reduces costs for developers by making input tokens three times cheaper and output tokens twice as cheap. Paying developers currently access this preview model, which also includes improved function calling for complex commands and a dedicated JSON mode to guarantee valid formatting.
The company introduces the Assistants API to help developers easily build sophisticated AI applications with built-in capabilities like code interpretation, information retrieval, and enhanced function calling. Alongside this developer tool, OpenAI allows ChatGPT users to create custom GPTs tailored for specific use cases without needing any coding skills. These customized bots aim to provide highly specialized experiences directly within the familiar ChatGPT interface.
OpenAI also brings notable updates to its GPT-3.5 Turbo model by expanding its default context window to 16K and improving its ability to follow formatting instructions by 38 percent. Furthermore, GPT-4 Turbo now possesses multimodal capabilities, meaning it processes and understands images in addition to text. These combined announcements represent a major leap forward in making AI tools more versatile, affordable, and accessible for both everyday users and software developers.