DeepSeek R1 (Jan '25)

About the DeepSeek R1 family

The open reasoning model that upended the industry in Jan 2025 — RL-trained chain-of-thought at ~$6M training cost, MIT-licensed, plus distilled versions from 1.5B to 70B that put reasoning on laptops. R1 1776 is Perplexity's decensored fine-tune.

How this one is configured

This configuration version/snapshot: Jan '25.

Who makes it

More from DeepSeek R1

• DeepSeek R1 0528 (May '25)
• DeepSeek R1 0528 Qwen3 8B
• DeepSeek R1 Distill Llama 70B
• DeepSeek R1 Distill Llama 8B
• DeepSeek R1 Distill Qwen 1.5B
• DeepSeek R1 Distill Qwen 14B
• DeepSeek R1 Distill Qwen 32B
• R1 1776