macOS
Homebrew已验证 · 100%
brew install yaplocal Homebrew formula metadata
安装
其他安装命令
Linux
Nix已验证 · 92%
nix profile install nixpkgs#yapnixpkgs package indexes · pkgs/by-name/ya/yap/package.nix · 来源: api.github.com
概览
软件包摘要
On-device audio transcription using Speech.framework
项目链接
命令和别名
- yap
历史
项目历史与用法
yap is a small macOS command-line transcription tool built around Apple's Speech.framework. It turns local audio, video, microphone input, or live system audio into text, subtitle formats, or JSON without sending the core speech-recognition work to a hosted service.
项目历史
The public repository was created on 2025-06-15, the same day the first GitHub release, 1.0.0, was published. Its README frames the project around macOS 26 and Apple's on-device Speech.framework, which places it in the wave of local-first media and AI-adjacent command-line tools that appeared as operating systems exposed better speech APIs.
By early 2026 the project had added real-time workflows beyond one-shot file transcription: the README documents `listen` for system audio, `dictate` for microphone input, and `listen-and-dictate` for mixed meeting-style transcription.
采用历史
yap's adoption story is still young and mostly package-manager/toolbox adoption rather than platform adoption. Its upstream README lists Homebrew and Mint installation, and the GitHub repository had more than a thousand stars by July 2026, enough signal for a niche macOS CLI but not enough history for a long institutional narrative.
使用方式
Users run `yap transcribe` or the shorter `yap <file>` style workflows to convert audio or video to text, SRT, WebVTT, or JSON. The README also shows package-nerd-friendly pipelines: downloading audio with yt-dlp, piping transcripts into an LLM, or producing caption files for video.
The live modes make it useful as a terminal-native dictation and meeting transcription tool, with macOS Screen Recording and Microphone permissions as the practical trust boundary.
为什么软件包爱好者会关心
yap is significant as a thin, scriptable wrapper over a native macOS speech-recognition stack. For package nerds, the interesting part is not a new speech model but the packaging of platform speech APIs into composable Unix-style commands and text/subtitle outputs.
时间线
- 2025-06-15: GitHub repository is created and release 1.0.0 is published.
- 2026-02-15: Release 1.2.0 is published.
- 2026: README documents file transcription, live system audio transcription, microphone dictation, and combined listen-and-dictate workflows.
Related projects
- Apple Speech.framework is the native macOS speech-recognition API used by yap.
- yt-dlp and llm appear in upstream examples as common pipeline companions.
来源
安全态势
风险级别：blue
broad file, network, media, or database tool signal.
风险分类器
blue 风险 · 中 置信度 · tool
原因
- broad file, network, media, or database tool signal
信号
- text:audio
安装行为
- 未记录 Homebrew bottle 元数据。
建议审查
在无人值守的代理使用前，请检查该工具是否读取明文凭据、写入远程状态、发布制品或调用插件。
可执行文件
已安装的可执行文件
| 命令 | 类型 | 暴露范围 | 备注 |
|---|---|---|---|
yap | 可执行文件 | 已索引可执行文件 | 从本地可执行文件索引发现。 |
新鲜度
版本和新鲜度
这些信号区分页生成时间、软件包管理器活动和上游发布比较。只有存在证据 URL 和可比较版本时，才会提示版本落后。
页面生成时间2026-08-03
管理器版本1.2.1
管理器更新时间2026-07-20
本地数据未知
上游不可用
检测到的最新版本未检测到
- OK没有生成新鲜度警告。
安装元数据
软件包元数据
| 软件包键 | brew:yap |
|---|---|
| 版本 | 1.2.1 |
| 软件包管理器 | Homebrew |
| 主页 | https://github.com/finnvoor/yap |
| 仓库 | https://github.com/finnvoor/yap |
| 最后更新 | 2026-07-20T11:07:32Z |
| Pulse | updated |
| Bottle | 未记录 |
| 服务 | 未声明 |
源数据库匹配
其他软件包管理器记录
匹配项来自外部软件包管理器索引，并与本地 Automic Vault 软件包链接分开显示。
Nix95%
yap
nix profile install nixpkgs#yap- normalized package name match
- 匹配方式：Yap
MacPorts95%
yap
sudo port install yap- normalized package name match
- 匹配方式：Yap
来源线索
由仓库数据生成
此页面由 av-web 从 scripts/generate-pkg-sqlite.py 生成的私有软件包 SQLite 工件提供。
Combined YAML source
View the package source record on GitHub.
使用的来源
- Geiger risk classifier
- cross-ecosystem install command graph
- curated package history
- external package-manager database matches
- pkg.so package database
- pkgdb category and tag curation