Groq has announced the launch of its DeepSeek R1 Distill Llama 70B model, which is now available for instant reasoning on its platform. The model boasts a processing speed of 275 teraflops per second (t/s), with expectations of further enhancements in speed as it is utilized. The DeepSeek R1 is designed for efficient reasoning tasks and is noted for its open-source capabilities. In addition to Groq, the DeepSeek R1 models are also being made available on platforms like Clarifai and Cloudflare Workers AI, expanding access to its advanced processing features. The model has undergone quantization, reducing its size from 720GB to 131GB, which allows for faster inference while requiring 160GB of VRAM. Users have expressed enthusiasm about the model's performance, highlighting its potential for various applications in AI-driven reasoning and coding tasks.