Artificial intelligence has been gradually creeping into daily life for years, making mundane tasks slightly more manageable. However, the advanced tools like ChatGPT and Claude often come with a catch – they require expensive subscriptions.
This changed when Chinese startup DeepSeek released its new open-source model, DeepSeek-R1, which can reason just as well as OpenAI’s flagship “reasoning” model, o1, but at a fraction of the cost. The company claims that R1 performs on par with o1 without consuming significantly more resources.
The implications are significant: an advanced AI model that’s free to the public and offers a bargain for developers who want to build apps on top of it. In contrast, big tech giants like OpenAI, Anthropic, Google, Meta, and Microsoft have collectively spent billions of dollars training their models.
DeepSeek says it invested less than $6 million in training its predecessor, DeepSeek-V3, paving the way for this groundbreaking development. The release of R1 opens up new possibilities for developers and challenges the dominance of big tech companies in the AI space.
Source: https://www.vox.com/technology/397330/deepseek-openai-chatgpt-gemini-nvidia-china