Introducing GPTS, and GPT Turbo Including Many Features
OpenAI has introduced GPTS, powerful language models that use deep learning to generate human-like text. This update is expected to be a significant advancement over the previous models.
Developers can link GPTs with reality.
In addition to using our built-in capabilities, you can define custom actions by making APIs available to GPT. Actions allow GPTs to integrate external data or interact with the real world, like plugins. They can be used to connect GPTs to databases, plug them into emails, or make them your shopping assistant. One example is integrating a travel listings database, connecting a user's email inbox, or facilitating e-commerce orders.
GPTs with privacy and safety in mind
ChatGPT ensures that your chats with GPTs are not shared. You can choose whether data can be sent to third-party APIs used by a GPT. When developers customize their GPT with actions or knowledge, the builder can choose if user chats with that GPT can be used to improve and train our models. These choices build upon users' existing privacy controls, including the option to opt your entire account out of model training.
GPT-4 Turbo Model
GPT-4 Turbo with 128K context and lower prices, the new Assistants API, GPT-4 Turbo with Vision, DALL·E 3 API, and more.
- New GPT-4 Turbo model supports 128K context window and is more affordable.
- New Assistants API simplifies building assistive AI apps with custom goals and tool integration.
- New features in the platform: vision, image generation (DALL·E 3), and text-to-speech.
GPT-4 Turbo has a 128k context window and knowledge of world events up to April 2023, allowing it to fit over 300 pages of text in a single prompt. The developer team optimized performance to offer GPT-4 Turbo at 3x cheaper input token price and 2x cheaper output token price compared to GPT-4.
GPT-4 Turbo is now available for developers to try by using gpt-4-1106-preview in the API. The stable production-ready model will be released in the upcoming weeks.
Function calling updates
Models can intelligently output JSON objects containing function arguments for app or external API functions described to the model.
We are pleased to inform you that GPT-4 Turbo has shown significant improvement in tasks requiring precise instruction, such as generating specific formats like XML. Additionally, GPT-4 Turbo now supports our new JSON mode.
Our first step towards helping developers build agent-like experiences within their own applications is the release of the Assistants API, which includes Retrieval and Code Interpreter functionalities.
An assistant is a purpose-built AI with specific instructions, leverages extra knowledge, and can call models and tools to perform tasks. The new Assistants API offers new functionalities, such as Code Interpreter and Retrieval, along with function calling to manage most of the complex tasks you used to perform manually. This enables you to construct high-quality AI applications with ease..
DALL·E 3
Developers can now integrate DALL·E 3 into their apps and products through our Images API by specifying dall-e-3 as the model we recently launched to ChatGPT Plus and Enterprise users. Companies like Snap, Coca-Cola, and Shutterstock have used DALL·E 3 to create images and designs for their clients and campaigns.
Lower prices and higher rate limits
OpenAI decreasing prices across the platform to pass on savings to developers (all prices below are expressed per 1,000 tokens):
GPT-4 Turbo input tokens are 3x cheaper than GPT-4 at $0.01 and output tokens are 2x cheaper at $0.03.
GPT-3.5 Turbo input tokens are 3x cheaper than the previous 16K model at $0.001 and output tokens are 2x cheaper at $0.002. Developers previously using GPT-3.5 Turbo 4K benefit from a 33% reduction on input tokens at $0.001. Those lower prices only apply to the new GPT-3.5 Turbo introduced today.
Fine-tuned GPT-3.5 Turbo 4K model input tokens are reduced by 4x at $0.003 and output tokens are 2.7x cheaper at $0.006. Fine-tuning also supports 16K context at the same price as 4K with the new GPT-3.5 Turbo model.
Model customization
GPT-4 fine-tuning experimental access
OpenAI creating an experimental access program for GPT-4 fine-tuning. Preliminary results indicate that GPT-4 fine-tuning requires more work to achieve meaningful improvements over the base model than the substantial gains realized with GPT-3.5.
Copyright Shield
OpenAI is committed to protecting our customers with built-in copyright safeguards in our systems. By introducing Copyright Shield—"we will now step in and defend our customers, and pay the costs incurred, if you face legal claims around copyright infringement." This applies to generally available features of ChatGPT Enterprise and our developer platform.