TaoShengTaoSheng
安卓语音生活记录助手:持续收音、转写和声纹区分,整理待办并按原话检索记忆。Android voice lifelog: captures speech, transcribes and identifies speakers, then organizes tasks and retrieves memories with source references.
开源软件;模型服务另计Open-source software; model services charged separately
功能Features
- ✓✓持续音频记录Continuous audio capture
语音检测后记录人声,可暂停与设置停录时段Voice activity detection captures speech; pause and recording schedules are available
- ✗✗屏幕历史记录Screen history
可选临时截图识字,识别后丢弃,不保留屏幕历史Optional temporary screenshot OCR; screenshots are discarded after recognition, not retained as screen history
- ✓✓说话人区分Speaker separation
本机声纹与多人分段;识别结果需核对Local voiceprints and diarization; verify identification results
- ✓✓记忆问答与回顾Memory chat & recaps
待办、每日回顾和带原话出处的问答Tasks, daily recaps and Q&A with transcript references
- 部分Partial本地处理选项Local processing options
本地转写加局域网自建整理模型;云服务需分别配置Local transcription plus a self-hosted LAN model; configure cloud services separately
- ✓✓Agent/API 集成Agent / API integration
Markdown/JSON 推送与自定义 WebhookMarkdown/JSON push and custom webhooks
价格与免费限制Price and free-tier limits
MIT 代码可免费使用;云转写、大模型、存储和推送服务可能另收费,模型与原生组件另有许可。MIT code is free to use; cloud transcription, AI, storage and push services may cost extra. Models and native components have separate licenses.
价格随地区、促销变化,以官网为准。Prices vary by region and promotion; official sites are authoritative.
免费档限制Free-tier limits
需 Android 10+、arm64;不是原生 Mac 应用。早期验证版,后台稳定性依机型。支持本地 SenseVoice 与局域网模型,但云转写/LLM、推送和云备份会发送所需数据;不能概括为任何配置都不出设备。录音前告知并取得所需同意,核对保存期限;未实测。Requires Android 10+ on arm64; not a native Mac app. Early validation release; background reliability varies by device. Local SenseVoice and LAN models are options, but cloud ASR/LLMs, push and backups send required data off-device. Do not assume all configurations stay local. Inform participants, obtain required consent and check retention; not tested.
概况Overview
软件资料采集日期 2026-11-14Software details collected 2026-11-14
GitHub 数据更新于 2026-11-16 19:57(北京时间)。GitHub data updated 2026-11-16 19:57 (UTC+8).