Introducing OpenAI o1
📰 OpenAI News
OpenAI introduces o1, a new series of AI models designed to spend more time thinking before responding, with improved reasoning capabilities in science, coding, and math
Action Steps
- Explore the OpenAI o1 documentation and technical research post to understand its capabilities and limitations
- Evaluate the potential applications of OpenAI o1 in various domains, such as science, coding, and math
- Develop and test AI models using the OpenAI o1 series to improve reasoning and problem-solving capabilities
- Integrate OpenAI o1 with other AI tools and frameworks to create more comprehensive AI solutions
Who Needs to Know This
AI researchers, data scientists, and software engineers can leverage OpenAI o1 to develop more advanced AI applications, while product managers and entrepreneurs can explore new use cases for this technology
Key Insight
💡 OpenAI o1 represents a significant advancement in AI capability, with improved reasoning and problem-solving capabilities, but may not yet have all the features of previous models like ChatGPT
Share This
🤖 OpenAI o1: a new series of AI models that think before they respond, with improved reasoning capabilities in science, coding, and math 🚀
Key Takeaways
OpenAI introduces o1, a new series of AI models designed to spend more time thinking before responding, with improved reasoning capabilities in science, coding, and math
Full Article
# Introducing OpenAI o1 | OpenAI
[Skip to main content](https://openai.com/index/introducing-openai-o1-preview#main)
[](https://openai.com/)
* [Research](https://openai.com/research/index/)
* Products
* [Business](https://openai.com/business/)
* [Developers](https://openai.com/api/)
* [Company](https://openai.com/about/)
* [Foundation(opens in a new window)](https://openaifoundation.org/)
[Try ChatGPT(opens in a new window)](https://chatgpt.com/?openaicom-did=69e5e09a-4a0a-43aa-b395-55f9d62fff1e&openaicom_referred=true)
* Research
* Products
* Business
* Developers
* Company
* [Foundation(opens in a new window)](https://openaifoundation.org/)
Introducing OpenAI o1 | OpenAI
Table of contents
* [How it works](https://openai.com/index/introducing-openai-o1-preview#how-it-works)
* [Safety](https://openai.com/index/introducing-openai-o1-preview#safety)
* [Whom it’s for](https://openai.com/index/introducing-openai-o1-preview#whom-its-for)
* [OpenAI o1-mini](https://openai.com/index/introducing-openai-o1-preview#openai-o1-mini)
* [How to use OpenAI o1](https://openai.com/index/introducing-openai-o1-preview#how-to-use-openai-o1)
* [What’s next](https://openai.com/index/introducing-openai-o1-preview#whats-next)
September 12, 2024
[Product](https://openai.com/news/product-releases/)
# Introducing OpenAI o1‑preview
Listen to article
Share
_Update on September 17, 2024: Rate limits are now 50 queries per week for o1‑preview and 50 queries per day for o1‑mini._
We've developed a new series of AI models designed to spend more time thinking before they respond. They can reason through complex tasks and solve harder problems than previous models in science, coding, and math.
Today, we are releasing the first of this series in ChatGPT and our API. This is a preview and we expect regular updates and improvements. Alongside this release, we’re also including [evaluations](https://openai.com/index/learning-to-reason-with-llms/) for the next update, currently in development.
## How it works
We trained these models to spend more time thinking through problems before they respond, much like a person would. Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes.
In our tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology. We also found that it excels in math and coding. In a qualifying exam for the International Mathematics Olympiad (IMO), GPT‑4o correctly solved only 13% of problems, while the reasoning model scored 83%. Their coding abilities were evaluated in contests and reached the 89th percentile in Codeforces competitions. You can read more about this in our [technical research post](https://openai.com/index/learning-to-reason-with-llms/).
As an early model, it doesn't yet have many of the features that make ChatGPT useful, like browsing the web for information and uploading files and images. For many common cases GPT‑4o will be more capable in the near term.
But for complex reasoning tasks this is a significant advancement and represents a new level of AI capability. Given this, we are resetting the counter back to 1 and naming this series OpenAI o1.
## Safety
As part of developing these new models, we have come up with a new safety training approach that harnesses their reasoning capabilities to make them adhere to safety and alignment guidelines. By being able to reason about our safety rules in context, it can apply them more effectively.
One way we measure safety is by testing how well our model continues to follow its safety rules if a user tries to bypass them (known as "jailbreaking"). On one of our hardest jailbreaking tests, GPT‑4o scored 22 (on a scale of 0-100) while our o1‑preview model scored 84. You can read more about this in the [system card](https://openai.com/index/openai-o1-system-card/) and our [research post](https://openai.c
[Skip to main content](https://openai.com/index/introducing-openai-o1-preview#main)
[](https://openai.com/)
* [Research](https://openai.com/research/index/)
* Products
* [Business](https://openai.com/business/)
* [Developers](https://openai.com/api/)
* [Company](https://openai.com/about/)
* [Foundation(opens in a new window)](https://openaifoundation.org/)
[Try ChatGPT(opens in a new window)](https://chatgpt.com/?openaicom-did=69e5e09a-4a0a-43aa-b395-55f9d62fff1e&openaicom_referred=true)
* Research
* Products
* Business
* Developers
* Company
* [Foundation(opens in a new window)](https://openaifoundation.org/)
Introducing OpenAI o1 | OpenAI
Table of contents
* [How it works](https://openai.com/index/introducing-openai-o1-preview#how-it-works)
* [Safety](https://openai.com/index/introducing-openai-o1-preview#safety)
* [Whom it’s for](https://openai.com/index/introducing-openai-o1-preview#whom-its-for)
* [OpenAI o1-mini](https://openai.com/index/introducing-openai-o1-preview#openai-o1-mini)
* [How to use OpenAI o1](https://openai.com/index/introducing-openai-o1-preview#how-to-use-openai-o1)
* [What’s next](https://openai.com/index/introducing-openai-o1-preview#whats-next)
September 12, 2024
[Product](https://openai.com/news/product-releases/)
# Introducing OpenAI o1‑preview
Listen to article
Share
_Update on September 17, 2024: Rate limits are now 50 queries per week for o1‑preview and 50 queries per day for o1‑mini._
We've developed a new series of AI models designed to spend more time thinking before they respond. They can reason through complex tasks and solve harder problems than previous models in science, coding, and math.
Today, we are releasing the first of this series in ChatGPT and our API. This is a preview and we expect regular updates and improvements. Alongside this release, we’re also including [evaluations](https://openai.com/index/learning-to-reason-with-llms/) for the next update, currently in development.
## How it works
We trained these models to spend more time thinking through problems before they respond, much like a person would. Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes.
In our tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology. We also found that it excels in math and coding. In a qualifying exam for the International Mathematics Olympiad (IMO), GPT‑4o correctly solved only 13% of problems, while the reasoning model scored 83%. Their coding abilities were evaluated in contests and reached the 89th percentile in Codeforces competitions. You can read more about this in our [technical research post](https://openai.com/index/learning-to-reason-with-llms/).
As an early model, it doesn't yet have many of the features that make ChatGPT useful, like browsing the web for information and uploading files and images. For many common cases GPT‑4o will be more capable in the near term.
But for complex reasoning tasks this is a significant advancement and represents a new level of AI capability. Given this, we are resetting the counter back to 1 and naming this series OpenAI o1.
## Safety
As part of developing these new models, we have come up with a new safety training approach that harnesses their reasoning capabilities to make them adhere to safety and alignment guidelines. By being able to reason about our safety rules in context, it can apply them more effectively.
One way we measure safety is by testing how well our model continues to follow its safety rules if a user tries to bypass them (known as "jailbreaking"). On one of our hardest jailbreaking tests, GPT‑4o scored 22 (on a scale of 0-100) while our o1‑preview model scored 84. You can read more about this in the [system card](https://openai.com/index/openai-o1-system-card/) and our [research post](https://openai.c
DeepCamp AI