DeepSeek's first generation reasoning models with comparable performance to OpenAI-o1
Want to make some of these yourself?
