Ollama
Get up and running with large language models locally.
- Star 数
- 183.2K
- Fork 数
- 18.3K
- 未关闭 Issue
- 4.3K
- 最近提交
- 2026-10-15
数据来自 GitHub API · 更新于 2026-10-15
Ollama 是什么？
Ollama downloads and runs open models such as Llama, Qwen, Gemma, Mistral and DeepSeek with one command and exposes a local REST API that many apps integrate with. It handles quantized model files, GPU offloading and model customization through Modelfiles.
主要功能
- One-command model download and chat
- Local REST API (OpenAI-compatible endpoint)
- Modelfiles for custom system prompts and parameters
- Vision and embedding models
系统要求
- 平台
- Windows · macOS · Linux · Docker
- 显卡
- Optional; NVIDIA, AMD and Apple Silicon accelerated
- 内存
- 8 GB for 7B models, 16 GB for 13B, 32 GB+ for larger
如何安装 Ollama
根据项目 README 整理。开源项目更新很快，请以 GitHub 页面上的最新说明为准。
macOS / Windows
- 1Download and run the installer from ollama.com.
- 2Run a model.
ollama run llama3.2
Linux
- 1Install with the official script.
curl -fsSL https://ollama.com/install.sh | sh - 2Run a model.
ollama run llama3.2
Docker
- 1Start the server (add --gpus=all for NVIDIA).
docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama docker exec -it ollama ollama run llama3.2
优势
- Easiest way to run local LLMs
- Huge app ecosystem
不足
- Less control than raw llama.cpp
Ollama 替代品对比
| 软件 | Star 数 | 许可证 | 平台 | 显存 |
|---|---|---|---|---|
| Ollama Get up and running with large language models locally. | 183.2K | MIT | — | |
| llama.cpp LLM inference in C/C++ — the engine behind most local AI apps. | 131.2K | MIT | — | |
| Jan Open-source ChatGPT-style desktop app that runs 100% offline. | 45K | AGPL-3.0 | — | |
| LocalAI Self-hosted, OpenAI-compatible API for text, images, audio and more. | 49.5K | MIT | — | |
| GPT4All Private, on-device chat with local document Q&A. | 77.4K | MIT | — | |
| vLLM High-throughput LLM serving engine with PagedAttention. | 93.8K | Apache-2.0 | — |
常见问题
Ollama 免费吗？
是的——Ollama 是开源软件，采用 MIT 许可证。商用前请确认软件及模型权重的许可条款。
Ollama 支持 Windows、Mac 和 Linux 吗？
Ollama 支持：Windows, macOS, Linux, Docker。
运行 Ollama 需要什么配置？
Optional; NVIDIA, AMD and Apple Silicon accelerated. RAM: 8 GB for 7B models, 16 GB for 13B, 32 GB+ for larger
如何安装 Ollama？
1. Download and run the installer from ollama.com. 2. Run a model.
Ollama 有哪些好的替代品？
常见替代品包括 llama.cpp, Jan, LocalAI, GPT4All，详见本页对比表。