OpenAI o1 is now out of preview in ChatGPT. What’s changed since the preview? A faster, more powerful reasoning model that’s better at coding, math & writing. o1 now also supports image uploads, allowing it to apply reasoning to visuals for more detailed & useful responses.
OpenAI o1 is more concise in its thinking, resulting in faster response times than o1-preview. Our testing shows that o1 outperforms o1-preview, reducing major errors on difficult real-world questions by 34%.
The updated OpenAI o1 system card builds on prior safety work, detailing robustness evals, red teaming insights, and safety improvements using Instruction Hierarchy. It maintains a "medium" risk rating based on testing with an expanded suite of evaluations, reflecting it is safe
Plus and Team users will have access to OpenAI o1 today through the model selector, replacing o1-preview. Enterprise and Edu users will have access in one week.
We’re still working on adding support for tools like web browsing and file upload into OpenAI o1 in ChatGPT. We're also working on making `o1` available in the API with support for function calling, developer messages, Structured Outputs, and vision—stay tuned.
Today we’re also adding ChatGPT Pro, a new plan that allows us to offer the best of our models & tools at scale, including unlimited access to OpenAI o1 and a Pro-only version of o1 that thinks longer for even more reliable responses. https://openai.com/index/intro...
@OpenAI I want to try the reindeerinforcement learning!!!
@OpenAI ngl making the first model preview and then the second one out of preview was kind of genius
@OpenAI Is there any timeline for full o1 via API?
@OpenAI >also working on making o1 available in the API with support for function calling, developer messages, Structured Outputs, and vision So we can roll our own o1 + code interpreter with function calling if need be!
@OpenAI I still see preview
@OpenAI log out and in to see it
@OpenAI This is so exciting, way to start the 12 days with a bang
@OpenAI Really impressed with this model. Here are my tests for those who are interested in knowing how to make sense of the capabilities of o1: https://youtu.be/1PWD21aeMx8
@OpenAI i’m crying
@OpenAI Would love to know the reason why that o1 is a better writer according to this post, yet evidence supported opposite for 4o preview. https://x.com/DeepwriterAI/sta...
@OpenAI Looks interesting but still waiting for the camera feature in the advanced voice mode.
@OpenAI Wonderful, this is so impressive.
@OpenAI This is good
@OpenAI Finally I can make my wife calculate how much she’s spending in stores
@OpenAI Well that's awkard..
@OpenAI Image understanding for o1 is great but inability to upload PDFs and other documents is disappointing.
@OpenAI Game-changer! What will you create first?
@OpenAI The obsession with optimizing for benchmarks in AI labs has led to a myopic focus on superficial solutions like fine-tuning, exemplified by the o1 series and test-time training. We need to address the root problems not self-congratulate over appearances. https://ai-cosmos.hashnode.dev...
@OpenAI The biggest brains on earth just a click away.
@OpenAI Impressive🔥 Are we expecting Sora soon?
@OpenAI I don't think you can be better than Claude Even Claude lobotomized itself Still June version better If a company wants to be number one bring a model that can output 32k context at once What is your output context size still lame 4k? 2k?
@OpenAI 😎 Awesome news! The new o1 sounds really powerful. Can't wait to see how it performs in coding, math and writing. 🤩 And image uploads? That's a great addition. Looking forward to more detailed and useful responses. 😃
@OpenAI whoa!
@OpenAI engineers bout to go out of business😂
@OpenAI Join my FREE TG here for Crypto Alphas/Airdrop guides https://t.me/geormanshub
@OpenAI OpenAI o1 System Card: This paper discusses the safety and capabilities of the OpenAI o1 model series, which are large language models trained using reinforcement learning to perform complex reasoning. The results showed that the o1 model series significantly outperformed the
@OpenAI How about you take the threat of an intelligence explosion seriously and start seriously pushing for a global non-proliferation treaty?