# DeepSeek

> Open-weight reasoning and chat models.

DeepSeek has a SharkScore of 73.8/100, ranked #2 of 8 in Models (updated 2026-11-17).

- Website: https://www.deepseek.com (deepseek.com)
- Kind: Model · Pricing: Open source
- Maker: DeepSeek
- Repository: https://github.com/deepseek-ai/DeepSeek-V3
- Categories: Models, Chat assistants
- Verified by the maker: no

## Why it scores 73.8

| Part | Score (0–100) | Weight |
|---|---|---|
| Adoption | 80 | 30% |
| User reviews | 85 | 25% |
| Maintenance | 77 | 20% |
| Trust & transparency | 80 | 15% |
| Momentum | 11 | 10% |

Evidence covers 100% of the score's weight. Compared with other listings in Models.

## Evidence

| Signal | Value |
|---|---|
| Monthly visits | — |
| GitHub stars | 100K |
| Stars, last 90 days | +2.5K |
| Downloads, 30 days | 6M |
| Reviews | 4.4 stars from 74 |
| Last update | 30 days ago |
| Uptime, 30 days | — |
| Public pricing / Free tier / API / MCP | yes / yes / yes / no |

## Top alternatives

1. [Qwen](https://rankshark.pro/tools/qwen) — SharkScore 79.5. Alibaba's open models for text, code, vision and audio.
2. [Llama](https://rankshark.pro/tools/llama) — SharkScore 64.6. Meta's family of open-weight language models.
3. [gpt-oss](https://rankshark.pro/tools/gpt-oss) — SharkScore 63.5. OpenAI's open-weight reasoning models.
4. [Ollama](https://rankshark.pro/tools/ollama) — SharkScore 83.9. Run open models on your own computer.
5. [Hugging Face](https://rankshark.pro/tools/hugging-face) — SharkScore 83.1. The home of open models, datasets and Spaces.

---
Source: https://rankshark.pro/tools/deepseek · Method v2026-11-17.1: https://rankshark.pro/how-ranking-works · Payment never changes a rank.
