DeepSeek-R1, a new large language model developed in China, is gaining attention as a cost-effective and open alternative to existing reasoning models like OpenAI's o1. Released on January 20, R1 has demonstrated performance on par with o1 in areas such as chemistry, mathematics, and coding, while its open-weight accessibility allows researchers to analyze and build upon its underlying algorithm. Cost-effective usage is highlighted as R1 is reportedly 30 times cheaper to operate compared to o1, potentially leading to greater adoption in research settings.
DeepSeek has released R1 as 'open-weight', allowing researchers to study and build upon the algorithm, which is uncommon for AI models, enhancing collaborative research.
The cost efficiency of R1 is significant; an experiment that would cost over £300 with OpenAI's model costs less than $10 with DeepSeek's R1, making it accessible.
Collection
[
|
...
]