{"repo":"14790897/tg-voice-ai","free":true,"listed":false,"github":"https://github.com/14790897/tg-voice-ai","clone":"git clone https://github.com/14790897/tg-voice-ai.git","description":"telegram voice chat ai bot with image generate","language":"TypeScript","stars":256,"topics":["ai","chat","telegram","voice"],"license":null,"category":"chat-messaging","readme_excerpt":"电报语音对话机器人 功能 发送语音或者文字，收到语音+文字的聊天回复 使用方法 我部署的 https://t.me/tg voice ai bot 可以直接使用 自己部署 1. 在 botfather 创建机器人 2. 部署 cf 程序 3. Configure the Cloudflare Worker secrets: - tg token : Telegram bot token from BotFather. - siliconflow token (optional): SiliconFlow API token for the SiliconFlow provider. - tts provider (optional): Choose workers , siliconflow , or deepgram to pick the TTS backend (defaults to siliconflow when a token is set, otherwise workers ). - tts lang (optional): Language code passed to MeloTTS when using the Workers provider (defaults to en ). - tts speaker (optional): Deepgram Aura speaker id (defaults to angus ). - tts encoding , tts container , tts sample rate , tts bit rate (optional): Additional Deepgram Aura audio options. 4. Visit the worker domain /init once to register the Telegram webhook (for example https://voiceai.14790897.xyz/init). 5. 与机器人聊天，支持语音和文字 6. 可以在代码中修改使用模型 演示 日志查看 pnpm dlx wrangler tail 语音模型选择 wrangler.json 中可以选择语音模型 tts provider","default_branch":null,"files":null,"tree":[],"storefront":"/r/14790897","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/14790897/tg-voice-ai/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}