## Metadata - Author: [[openai.com]] - Full Title:: New Models and Developer Products Announced at DevDay - Category:: #🗞️Articles - Document Tags:: [[LLMs]], - URL:: https://openai.com/blog/new-models-and-developer-products-announced-at-devday - Read date:: [[2023-11-07]] ## Highlights > GPT-4 Turbo is more capable and has knowledge of world events up to April 2023 ([View Highlight](https://read.readwise.io/read/01hek7p4tv5tvpehp167xdbyne)) > it can fit the equivalent of more than 300 pages of text in a single prompt. ([View Highlight](https://read.readwise.io/read/01hek7pcsg528j2sfeyfj38ztw)) > The new `seed` parameter enables **reproducible outputs** by making the model return consistent completions most of the time. ([View Highlight](https://read.readwise.io/read/01hek7v81qk81tzgsvp4dm8fqv)) > **persistent and infinitely long threads**, which allow developers to hand off thread state management to OpenAI and work around context window constraints. With the Assistants API, you simply add each new message to an existing `thread`. ([View Highlight](https://read.readwise.io/read/01hek7z36vmwzg6387j88prq74)) > **Retrieval**: augments the assistant with knowledge from outside our models, such as proprietary domain data, product information or documents provided by your users. This means you don’t need to compute and store embeddings for your documents, or implement chunking and search algorithms. The Assistants API optimizes what retrieval technique to use based on our experience building knowledge retrieval in ChatGPT. ([View Highlight](https://read.readwise.io/read/01hek81zev3f5fc900vktyr621)) > GPT-4 Turbo can accept images as inputs in the Chat Completions API ([View Highlight](https://read.readwise.io/read/01hek8754vx89xqxep3n60xy99)) > Companies like Snap, Coca-Cola, and Shutterstock have used DALL·E 3 to programmatically generate images and designs for their customers and campaigns ([View Highlight](https://read.readwise.io/read/01hek87vnq29fnh1kzwq715gfp))