Send "文字转语音 <text>" and the bot replies with bibibobo speech.
Bibibobo TTS
Turn a short command into speech: send 文字转语音 空气 and the bot replies with a voice
message that literally says the two characters 空气.
The synthesizer is a Python port of the bibibobo engine (a formant/"babble" synth — one sawtooth oscillator, two band-pass filters, one syllable per character, character hash picks the pitch). The port is byte-identical to the original JS implementation at 44.1 kHz / 16-bit mono WAV. No cloud TTS, no API key, no recording.
Usage
| You send | The bot replies with |
|---|---|
文字转语音 空气 | voice saying 「空气」 |
文字转语音 体验 | voice saying 「体验」 |
文字转语音 | usage hint |
Only the text after the trigger phrase is synthesized — nothing else. The trigger phrase is configurable; the default pipeline reply is suppressed so the model does not answer twice.
Configuration
| Name | Type | Default | Description |
|---|---|---|---|
trigger | string | 文字转语音 | Message must start with this |
step | float | 0 | Seconds per syllable; 0 = default 0.13 |
send_mode | select | voice | voice 语音消息 / file WAV 文件 |
max_chars | integer | 120 | Longer input is truncated |
Notes
- Punctuation rules are kept:
?raises the tail,!raises the whole sentence,…lowers the sentence and slides the tail down. - If the platform adapter cannot send base64 voice, switch
send_modetofile. - Synthesis runs in the plugin subprocess; a 4-second clip takes roughly 0.2 s on CPU.
Development
pip install -U langbot_plugin
cp .env.example .env # fill DEBUG_RUNTIME_WS_URL / PLUGIN_DEBUG_KEY from WebUI
lbp run
lbp build # dist/<author>-<name>-<version>.lbpkg
License of the synthesis algorithm: MIT (路墨小剧场 / 哈基科技, Lumoswitch).
Use Bibibobo TTS in Feishu/Lark bots, DingTalk bots, WeCom bots, WeChat bots, QQ bots, Slack bots, Discord bots, Telegram bots, LINE Bots, and Matrix Bots.
Host your bot on LangBot Cloud and connect this extension to the collaboration platforms your team already uses.