OpenAI Hosts First Developer Conference: A Summary of Key Announcements

OpenAI recently held its inaugural developer event, a day filled with exciting announcements and product launches. If you missed it, here's a comprehensive summary of all the major developments.

Milestones for ChatGPT

A year after its initial release, ChatGPT has seen significant adoption. OpenAI reported over 100 million weekly users of the tool, making it one of the fastest consumer products to reach the 100 million monthly user mark.

The company also highlighted that over 2 million developers are leveraging its API to create innovative solutions.

Introduction of GPT-4 Turbo

OpenAI unveiled an enhanced version of the popular GPT-4 model, named GPT-4 Turbo, at the event. The new Turbo model is available in two versions: one for text analysis only and another for understanding both text and images.

GPT-4 Turbo boasts a context window of 128,000 tokens (approximately 100,000 words), four times larger than its predecessor. It also extends the knowledge cut-off to April 2023, compared to GPT-4's September 2021 cut-off.

⚡ Learn more about GPT-4 Turbo

Custom GPT Models for Users

In a major announcement, OpenAI revealed plans to allow users to create their own versions of GPT for various use cases. Users can build these bots using prompts, without any coding knowledge. Furthermore, enterprise customers can construct internal-only GPTs based on the company's knowledge base.

Developers can connect these GPTs to databases or knowledge bases, such as emails, to import external information.

Launch of GPT Store for User-Created AI Bots

OpenAI plans to enable users to publish their GPTs to a store, set to open later this month. Initially, the store will feature creations from "verified builders."

CEO Sam Altman discussed potential compensation for creators of popular GPTs.

New Assistants API for Developers

OpenAI introduced a new Assistants API, allowing developers to create their own "agent-like experiences." These agents can retrieve external knowledge or call programming functions for specific actions, with use cases ranging from coding assistants to AI-powered vacation planners.

Release of DALL-E 3 API

The company's text-to-image model, DALL-E 3, is now accessible via API with built-in moderation tools. OpenAI has priced this model at $0.04 per generated image.

⚡ Learn more about DALL-E 3

Launch of New Text-to-Speech APIs

OpenAI launched a new text-to-speech API, the Audio API, featuring six preset voices: Alloy, Echo, Fable, Onyx, Nova, and Shimer. This API provides access to two generative AI models.

Copyright Shield Program for Businesses

OpenAI announced a new program, Copyright Shield, promising to protect businesses using the company's products from copyright claims. The company pledged to cover legal fees for customers facing IP lawsuits over content created by OpenAI's tools.

Other Notable Announcements

• OpenAI also announced a program to assist companies in building their customer models with the help of the company's researchers.
• The model picker in ChatGPT will be removed to simplify the user experience.
• The company also launched the next version of its open-sourced speech recognition model, Whisper large v3.
• Doubling of the tokens per limit rate for all paying GPT-4 customers.