A customized, self-hosted server for your Folotoys.
- Speech in and voice out
- LLM(Large Language Model) supports OpenAI/Azure OpenAI/Google Gemini/Baidu Qianfan/Dify/Moonshot
- SST(Sound to Text) supoorts OpenAI Whisper/Azure Speech Service/Azure Whisper/Aliyun ASR/Dify
- TTS(Text to Sound) supports OpenAI TTS/Azure Speech Service/Elevenlabs/Edge TTS/Aliyun TTS/Netease EmotiVoice/OpenVoice
- AI Conversational data is published to a MQTT Broker
The documentation is available at docs.folotoy.com.
For an opt-in language-following role and fixed-language voices, see the multilingual role example.
For a streamlined deployment experience, you can use the folotoy-server-deploy OpenClaw Skill. This skill automates the setup using Docker Compose and allows easy configuration of your STT, LLM, and TTS API keys.
To use this skill, make sure you have OpenClaw installed. You can then import the skill (folotoy-server-deploy.skill) and follow its instructions to deploy the FoloToy Server.
Copy-paste the below code and press enter
git clone https://github.com/FoloToy/folotoy-server-self-hosting.git folotoy-server