OpenAI announces general availability of its GPT-4 API: Here’s what it means

OpenAI has announced the general availability of GPT-4, its latest text-generating model. Developers can now access GPT-4 through its API.

Currently, GPT-4 API access will be limited to developers with a history of successful payments, but the company has planned to open access to new developers by the end of this month. OpenAI will also increase the availability depending upon the compute availability. 

“All existing API developers with a history of successful payments can access the GPT-4 API with 8K context,” OpenAI said in a statement. 

GPT-4 is a successor of GPT-3.5 and is much more capable than its predecessor. It can generate text including codes and can accept image and text inputs. It is trained using publicly available data, which includes data from public web pages and data that OpenAI is licensed to.  

“Millions of developers have requested access to the GPT-4 API since March, and the range of innovative products leveraging GPT-4 is growing every day,” the company said. 

OpenAI is also planning to allow developers to fine-tune GPT-4 and GPT-3.5 Turbo later this year, with their own data. This capability is currently available to developers for other OpenAI’s text-generating models. 

“We are working on safely enabling fine-tuning for GPT-4 and GPT-3.5 Turbo and expect this feature to be available later this year,” OpenAI said. 

What else?

OpenAI has also announced that it is making its DALL-E 2 and Whisper APIs generally available. DALL-E 2 and Whisper are image-generating and speech-to-text models from OpenAI. 

In addition to this, OpenAI has announced that it will be deprecating some of its older models using Completions API. The company has taken this decision to increase investment in Chat Completion API and to optimise its compute capacity. While the models based on Completions API will be available for the next six months, the company will label them as legacy in its developer documentation. 

For the unversed, the Completions API endpoint provides the completion for a single prompt and takes a single string as input whereas the Chat Completions API endpoint provides the responses for a given dialog and requires the input in a specific format corresponding to the message history.

“We plan for future model and product improvements to focus on the Chat Completions API, and do not have plans to publicly release new models using the Completions API,” the company said. 

Starting January 4, 2024, the old completion models will no longer be available and developers using these models will have to manually upgrade their integrations by January 4. If they want to continue using fine-tune old models beyond the deadline, they have to “fine-tune replacements atop the new base GPT-3 models”.

The post OpenAI announces general availability of its GPT-4 API: Here’s what it means appeared first on Techlusive.