README / README.md
gxcsoccer's picture
org card: download link to deskmind-ai/deskmind releases
0889e40 verified
|
Raw History Blame Contribute Delete
2.87 kB
---
title: README
emoji: 🟧
colorFrom: gray
colorTo: red
sdk: static
pinned: false
short_description: Open-source computer use for your Mac
---
![DeskMind 得心 — 得心,应手。](https://raw.githubusercontent.com/deskmind-ai/deskmind/main/brand/social/README-bilingual-light.svg)
**Small enough to run on your Mac. Smart enough to ask.**
小到能在你的 Mac 上跑,聪明到知道该问你。
[Website · 官网](https://deskmind.dev) · [Download for Mac](https://github.com/deskmind-ai/deskmind/releases/latest) ·
[GitHub](https://github.com/deskmind-ai) · [Bench](https://github.com/deskmind-ai/bench)
DeskMind is open-source computer use for your Mac. It reads the screen, works out the next step and acts, with small
models running locally with MLX. When a task could mean two things, it asks instead of guessing.
| Model | What it does |
|---|---|
| [brain-0.8b](https://huggingface.co/deskmind/brain-0.8b) | Fast tier: decides each step, with a probability for every option (MLX, 8-bit) |
| [brain-4b](https://huggingface.co/deskmind/brain-4b) | Strong tier: checks the steps the 0.8B is unsure of (MLX, 8-bit) |
| [eyes-4b](https://huggingface.co/deskmind/eyes-4b) | Screen grounding: finds the target on a screenshot, for apps without an accessibility tree |
**Current release: G18b** (revision `g18b-q8`, also `main`). On the real macOS desktop (bench v25, 13 tasks × 3 runs
through the DeskMind app on an M4 Pro), the 0.8B → 4B router at threshold 0.96 passed **39 of 39** runs and never said
"done" early. A decision takes about 0.5 s when the 0.8B answers and about 3.6 s when the 4B checks. Small sample;
what it still gets wrong is in [brain results](https://github.com/deskmind-ai/brain/blob/main/docs/results.md). Any
agent can use the planner through `/v1/systemone`.
---
**DeskMind 得心**是在你 Mac 上运行的开源电脑操作 AI:看屏幕、想下一步、动手操作,靠的是本机的小模型(MLX);任务有两种理解时,
它先问你,不瞎猜。
- **brain-0.8b**:判断每一步,每个选项都给出概率;**brain-4b**:复核 0.8B 没把握的步骤;**eyes-4b**:在截图上找到要点的位置,
用于没有辅助功能结构的应用。
- **当前版本 G18b**:真实 macOS 桌面上(bench v25,13 个任务各跑 3 次,经 App 运行,M4 Pro),0.8B → 4B 路由(门槛 0.96)
39 次全部通过,没有一次没做完就说完成。样本不大,还做不好的地方见
[brain results](https://github.com/deskmind-ai/brain/blob/main/docs/results.zh-CN.md)。
- 国内下载:ModelScope 上有同样的文件([gxcsoccer/brain-0.8b](https://www.modelscope.cn/models/gxcsoccer/brain-0.8b)、
[gxcsoccer/brain-4b](https://www.modelscope.cn/models/gxcsoccer/brain-4b)、
[gxcsoccer/eyes-4b](https://www.modelscope.cn/models/gxcsoccer/eyes-4b))。