This Chinese AI model can take over OpenAI’s o1 in advanced reasoning

A Chinese artificial intelligence (AI) model was released on Wednesday, claiming to adopt OpenAI’s o1 AI model in terms of advanced reasoning. The large language model (LLM), called DeepSeek-R1-Lite-Preview, is said to have outperformed the o1 model on several benchmarks. Notably, the AI ​​model can be tested for free on the internet, although the advanced reasoning function can only be used a limited number of times. Moreover, the AI ​​model also provides a transparent thought process that allows users to see how the output decision was made.

DeepSeek-R1 AI Model Revealed

Advanced reasoning is a relatively new ability for LLMs, allowing them to make decisions using multi-step thought processes. There are several advantages to this. First, such AI models can answer more complex questions and require insight into the deeper context and expert-level knowledge of the subject. Alternatively, such AI models could also fact-check themselves, minimizing the risk of hallucinations.

However, so far not many basic models are capable of advanced reasoning. Although some Mix-of-Agent (MoE) models can do this, they are made up of several smaller models. In the mainstream space, OpenAI o1 series models are known for this capability.

But on Wednesday, DeepSeek, a Chinese AI company, posted on X (formerly known as Twitter) announcing the release of the DeepSeek-R1-Lite-Preview model. The company claims it can outperform the o1 preview model on the AIME and MATH benchmarks. Notably, both of these test the mathematical and reasoning skills of an LLM.

Gadgets 360 employees gained access to the chatbot and discovered that the AI ​​model also shows the entire train of thought after submitting a question. This allows users to understand the logical connection made by the model and notice any shortcomings. During our testing, we found that the AI ​​model can answer complex questions.

The response time was also short, which made the conversation efficient. Currently, users only receive 50 messages to try out the “Deep Think” mode, which shows the model’s thinking process. Furthermore, this is currently the only free-to-use AI model with advanced reasoning. Interested parties can try out the AI ​​chatbot on the internet here.

Notably, the company has claimed that it will open the full version of the DeepSeek-R1 AI model in the near future, which would be a first for an LLM of this class.