一只住在闲置 iPhone / iPad 里的陪伴型桌宠。
写邮件给我，通常几天内回复：
App 里也可以直接发：设置 → 高级设置与支持 → 反馈。那样邮件会自动带上版本号和机型，不含任何对话内容。
能。它会眨眼、看你、随时间犯困或来精神，也会用自带的台词回应你。只是聊不深——想让它真正听懂你，需要接一个大模型服务。
设置 → AI 与记忆 → AI 服务，选一家服务商（OpenAI、Anthropic，或者局域网里自己跑的 Ollama），填入你自己的 API Key。费用由你和服务商结算，momo 不收费、不抽成、不中转。
先看屏幕上有没有一行小字提示，比如 Key 无效或余额不足。其次，它夜里本来就比较安静。设置 → 高级设置与支持 → 自检 可以看到每一路传感器和模型连接的实时状态。
默认不会。设置 → 声音与互动 →「主动开口」打开后，你在的时候它大约一到两小时会主动找你说一句，夜里不说，你一直不理它它会自己变少。
一个在设备上运行的语音模型，比系统音色自然得多。它不随 App 一起下载，需要时在 设置 → 声音与互动 里下载一次（约 200 MB，建议连 Wi-Fi）。
摄像头画面永不离开设备。完整说明见 隐私政策。
A companion desk pet that lives in a spare iPhone or iPad.
Email me. I usually reply within a few days:
You can also send it from inside the app: Settings → Advanced & Support → Feedback. That email pre-fills the version and device model, with no conversation content.
Yes. It blinks, watches you, gets sleepy or perks up over the day, and answers with its own built-in lines. It just can't go deep. To have it really understand you, connect a language model.
Settings → AI & Memory → AI providers. Pick a provider (OpenAI, Anthropic, or an Ollama server on your own network) and enter your own API key. You pay the provider directly; momo charges nothing, takes no cut, and never proxies your conversations.
First check for a small line of text on screen, such as an invalid key or no credit left. It's also naturally quieter at night. Settings → Advanced & Support → Self-check shows the live status of every sensor and the model connection.
Not by default. Turn on Settings → Voice & Interaction → "Speak on its own" and it will occasionally start a conversation, roughly once every one to two hours while you're around, never at night, and less often if you keep ignoring it.
An on-device speech model that sounds far more natural than the system voice. It isn't bundled with the app; download it once from Settings → Voice & Interaction when you want it (about 200 MB, Wi-Fi recommended).
Camera frames never leave your device. Full details in the Privacy Policy.