📋 Info
| GitHub Stars | ⭐ 92.0k Stars |
| License | Apache-2.0 |
| Language | Python |
| Version | R1-0528 |
| Updated | 2026-03-15 |
📖 Overview
DeepSeek-R1 is an inference-enhanced large language model developed by DeepSeek (92k Stars) that has reached world-class performance in code generation and mathematical reasoning. It utilizes a reinforcement learning + thought-chain training approach, granting it exceptional reasoning capabilities. Available in various sizes ranging from 1.5B to 32B parameters, it can run on consumer-grade GPUs. Its use is explicitly permitted for commercial purposes under the Apache 2.0 license. It supports context lengths of up to 128K tokens. It can be launched with just one click using Ollama. Currently, it is the most popular open-source inference model among developers in China.
✨ Features
- Outstanding capabilities in code and mathematical inference.
- Available in 1.5B–32B parameter sizes (compatible with consumer-grade GPUs).
- Apache 2.0 —— Explicitly licensed for commercial use
- 128K long context length
- Open-source large language models with excellent Chinese comprehension capabilities.
Advertisement