OpenAI is levelling the playing field for AI.
OpenAI has announced the launch of GPT-4o Mini, a smaller and significantly more affordable version of its powerful language model.
This new offering aims to empower startups and developers who had been previously priced out of using larger AI models, a significant barrier which led some to consider cheaper alternatives like Google’s Gemini 1.5 Flash. Recognizing the need for greater accessibility and the desire to target a wider customer base, OpenAI is now offering its users a more budget-friendly option.
“Say hello to GPT-4o, our new flagship model which can reason across audio, vision, and text in real time,” they wrote.
View this post on Instagram
Capabilities:
GPT-4o Mini boasts greater capabilities than GPT-3.5 despite remaining cheaper to operate. While it is designed for simpler tasks, its performance is impressive, achieving 82% score on the MMLU benchmark, surpassing GPT-3.5’s 70% score and rivalling the scores of competing light models like Claude 3 Haiku and Gemini 1.5 Flash.
OpenAI showcased the potential of GPT-4o Mini through partnerships with the fintech startup Ramp, who leveraged the tool to build a receipt-scanning tool that extracts expense data from images and the email client Superhuman, who used it to create an AI-powered email response suggestion feature.
GPT-4o Mini, available to ChatGPT users on Free, Plus, and Team plans starting today. It will also be available to developers via OpenAI’s API, offering support for text and vision processing, with plans to expand to multimodal inputs and outputs like video and audio in the near future.
Response:
Social media reactions to OpenAI’s latest model have been mixed and contradictory: while many are impressed with its unique features, others have pointed out the dangers of AI.
As a longtime voiceover guy, I am quite impressed over the nuances and subtleties of the new interactive audio GPT voices. Definitely a big upgrade,” one user praised.
“Cool but impractical,” another observed.
“Do you guys seriously not see the dangers in that? Being able to copy each and every voice and emotion?” another argued.
These range of reactions embody the controversial reception of AI, which has been embraced by a variety of tech giants and tech enthusiasts across the world despite the rising worries about its potentially cumbersome impact.
