DeepSeek R1
An open-source reasoning in LLM from DeepSeek!
🛡️ AgentReady threat assessment
MAESTRO 7-layer threat model + OWASP AIVSS risk score for DeepSeek R1, derived from its capabilities.
AIVSS 5.7 · Medium
View MAESTRO 7-layer threat model →Overview
deepseek r1(deepseek-r1.com) is an open-source AI reasoning model developed by DeepSeek AI, designed to offer advanced reasoning capabilities comparable to proprietary models. It provides developers with tools and resources for integrating AI into applications, emphasizing transparency and accessibility.
Key features
- 🧠 RL-Driven Reasoning: DeepSeek R1 pioneers a unique approach, applying reinforcement learning directly to the base model without prior supervised fine-tuning.
- 🚀 Powerful Architecture: Features a robust 671B parameter MoE architecture with 37B activated.
- 🔥 High-Performing Distilled Models: Including a Qwen-32B variant that outperforms OpenAI-o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.
- ✅ Open Source: DeepSeek has generously open-sourced both the main model and several smaller distilled models.
Use cases
- deepseek r1(deepseek-r1.com) is an open-source AI reasoning model developed by DeepSeek AI, designed to offer advanced reasoning capabilities comparable to proprietary models. It provides developers with tools and resources for integrating AI into applications, emphasizing transparency and accessibility.