Chinese fully open source reasoning model Deepseek -R1 at par with o1 at 20x cheaper price. This is big !

• DeepSeek's new reasoning model, R1, outperforms OpenAI’s o1 in benchmarks like AIME, MATH-500, and SWE-bench Verified.
• Available on Hugging Face under an MIT license, R1 offers distilled versions and is more cost-effective than OpenA...

For the majority of Indian startups : You cannot build a better datacenter than Microsoft. You cannot fundraise more than Sam or Elon. You cannot get more data than Google. But you actually can write 10x better code than any of them.

• Stanford and UW researchers created a $50 AI reasoning model, s1, rivalling OpenAI’s o1 and DeepSeek’s R1.
• Trained using Google's Gemini 2.0 Flash Thinking Experimental and distillation, s1 is open-sourced on GitHub.
• The achievemen...

• Just a week after DeepSeek's R1 AI model release, Hugging Face researchers, led by Leandro von Werra, announced Open-R1 to replicate and open-source R1's components.
• DeepSeek's R1, despite its performance, lacks transparency, pushing...