Check out Muse Spark, our first milestone in the quest for personal superintelligence! Scaling this with the team has been a total blast. Give it a spin and let us know what you think! 🥑
We released o3-mini today! Everyone can use it for free. It reasons hard, reasons fast, searches the web, and most importantly, knows research. Ask the model hard questions and brainstorm with it!
thank you to the external safety researchers who tested o3-mini.
we have now finalized a version and are beginning the release process; planning to ship in ~a couple of weeks.
also, we heard the feedback: will launch api and chatgpt at the same time!
(it's very good.)
Excited to present o3-mini today at @OpenAI 12 days with @sama, @markchen90 and @GregKamradt!
Check out the live demo where I asked the model to write a script to automatically evaluate itself on GPQA dataset from an interactive code generation/execution UI created by itself.
We also shared evals on Open AI o3-mini — a faster, distilled version of o3 which is optimized for coding, and the first version of o3 we expect to make available for use in early 2025.
we are going to take a little more time with our open-weights model, i.e. expect it later this summer but not june.
our research team did something unexpected and quite amazing and we think it will be very very worth the wait, but needs a bit longer.
Thrilled to release o1-mini, a model near and dear to my heart 💙. o1-mini is an efficient model in the o1 series that’s super performant in STEM reasoning, especially math and coding. I can’t wait to see what you all build with o1-mini!!
change of plans: we are going to release o3 and o4-mini after all, probably in a couple of weeks, and then do GPT-5 in a few months.
there are a bunch of reasons for this, but the most exciting one is that we are going to be able to make GPT-5 much better than we originally
Excited to announce 2nd Stanford Graph Learning Workshop on Wed Sept 28th with leaders from academia and industry to showcase recent advances of Graph Representation Learning across a wide range of applications. Program & free registration: snap.stanford.edu/graphlearning-…
I'm at #NeurIPS2024 this week. Excited to meet old and new friends. Our team at @OpenAI is hiring! If you are interested in (1) the most effective ways to compress intelligence that lead to product like 4o-mini, o1-mini, and more, (2) unlock agentic use cases with a mixture of
Very excited to finally share OpenAI's "deep research" model, which achieves twice the score of o3-mini on Humanity's Last Exam, and can even perform some tasks that would take PhD experts 10+ hours to do!
A few thoughts on the implications: Deep research can be seen as a new
