All your product insights in Kraftful are now powered by the o3-mini-high model.
What’s o3-mini-high, you ask?
It’s the latest and greatest reasoning model—a smaller version of the yet unreleased o3 model that reportedly outperformed humans on the ARC-AGI benchmark in December. “High” refers to the highest reasoning level we’ve chosen to run in Kraftful. While neither o3 nor o3-mini-high is AGI (yet), they’re very smart.
In our testing, o3-mini-high outshines o1 at grouping related insights, crafting concise yet context-rich summaries, and mapping insights back to original comments.
Seeing the improvements, our team dropped everything on Friday to get this live for you fast:
12:30 PM - OpenAI drops o3-mini
1:30 PM - “This could improve our analysis”
3:30 PM - o3-mini-high thoroughly tested & live in Kraftful
And that, kids, is how you ship on Fridays .