OpenAI to Reportedly Release the New Strawberry AI Model
in the Next 2 Weeks. Don't let the simple name fool you--the model could be capable of solving complex problems by thinking in multiple steps.
By Ben Sherry, Staff reporter
Artificial intelligence market leader OpenAI is reportedly preparing to launch its latest AI model, codenamed Strawberry, on ChatGPT in the next two weeks.
According to new reporting from The Information, OpenAI is planning to imminently launch Strawberry, a model capable of solving complex problems that have beguiled earlier models by thinking in multiple steps.
By all accounts, Strawberry will be considered a reasoning model, meaning that it will be capable of taking on more complicated requests that require multiple steps to complete, from solving tricky algebra problems to developing a comprehensive monthslong marketing campaign. The Information also reported that at launch, Strawberry will only be able to process text, unlike GPT-4o, OpenAI's flagship model that can process images and audio.
While this means Strawberry will reportedly be less prone to errors and hallucinations than GPT-4o, the trade-off is that the model is said to be slower. According to The Information, it can take anywhere from 10 to 20 seconds to respond to a query, and while the model is supposed to avoid deep thinking for simple requests, it doesn't always work that way in practice. Some people who had tested early versions of the model told The Information that these slightly better answers weren't worth the wait.
It's likely that due to its advanced capabilities, Strawberry will be more compute-heavy than its siblings and cost more to use. The Information reported that Strawberry "will likely have rate limits restricting users to some maximum number of messages per hour, with the potential for a higher-priced tier that's faster to respond."
Photo Credit: Getty Images.