OpenAI announced that it has applied a brand new model of its GPT-4o massive language mannequin to drive its ChatGPT chatbot, however it has declined to specify precisely how the up to date mannequin differs from its predecessor.
“To be clear, this is an improvement to GPT-4o and not a new frontier model,” the corporate posted on X (previously Twitter) Monday.
there's a brand new GPT-4o mannequin out in ChatGPT since final week. hope you all are having fun with it and test it out for those who haven't! we predict you'll prefer it 😃
— ChatGPT (@ChatGPTapp) August 12, 2024
After permitting customers to take their greatest guesses at defining the brand new iteration, which OpenAI is asking ChatGPT-4o-latest, the corporate added a few scant details on its Models page on Wednesday.
Described as a “dynamic model continuously updated to the current version of GPT-4o in ChatGPT,” GPT-4o-latest has a information cutoff of October 2023, and may accommodate 128,000 tokens, or 96,000 phrases, per dialog, simply because the earlier GPT-40 model did. It can output 16,384 tokens, or 12,288 phrases, on par with the newer GPT-40-mini mannequin and roughly quadruple what the older GPT-4o can do.
Unfortunately, onerous stats like these don’t present perception into what the brand new mannequin is definitely able to, and, apparently, neither will its builders.