The AI industry rarely made long-lasting hypes beyond its own domain. OpenAI changed that.
When OpenAI introduced ChatGPT AI, the company has successfully positioned itself at the vanguard of AI development. People were wowed, and rivals went scrambling for solutions.
And when OpenAI introduced GPT-4, users couldn't be happier.
This time, OpenAI is opening access to the API for GPT-4, and developers are the ones who are smiling wide.
In a blog post, OpenAI said that:
"We envision a future where chat-based models can support any use case."
GPT-4 can generate text (including code), and accept image and text inputs.
It's a huge improvement over GPT-3.5, its predecessor, which only accepted text, and performs at "human level" interactions on various professional and academic benchmarks.
By announcing the general availability of GPT-4 API to developers who have "history of successful payments," OpenAI is widening the reach of the sophisticated capabilities of GPT-4.
With, OpenAI dreams a future where "chat-based models can support any use case."
In the end, the company plans to open up the API access to new developers by the end of this July, and then start raising availability limits after that "depending on compute availability."
Back in March, OpenAI also launched its Chat Completions API so developers could "build conversational experiences and a broad range of completion tasks." At this time, Chat Completions takes up "97% of our API GPT usage."
On January 2024, the company plans to retire some of its older models that use the older Completions API, specifically GPT-3 and its derivatives, and replace them with the new "base GPT-3" models.
In anticipation for the switch, the company has urged developers to switch to Chat Completions.
Developers using the old models will have to manually upgrade their integrations by January 4, 2024, and those who wish to continue using fine-tuned old models beyond January 4 are required to fine-tune replacements atop the new base GPT-3 models.
"We will be providing support to users who previously fine-tuned models to make this transition as smooth as possible," OpenAI wrote. "In the coming weeks, we will reach out to developers who have recently used these older models, and will provide more information once the new completion models are ready for early testing."
GPT-4 API is now available to all paying OpenAI API customers. GPT-3.5 Turbo, DALL·E, and Whisper APIs are also now generally available, and we’re announcing a deprecation plan for some of our older models, which will retire beginning of 2024: https://t.co/H11w4IalNZ
— OpenAI (@OpenAI) July 6, 2023
It's worth noting though, while GPT-4 is a lot more powerful that GPT-3.5, and that it represents a significant milestone in generative AI models, it’s important to underscore that it’s still with its former self's limitations.
The model occasionally "hallucinates" facts, making reasoning errors, sometimes with confidence.
On top of that, OpenAI also has more goodies for developers.
Besides opening access to its GPT-4 model to paying customers, OpenAI also has other offerings that include GPT-3.5 Turbo, DALL·E, and Whisper have also been made generally available.