Memoh - Containerized AI Agent Platform
Containerized AI agent platform where each bot runs in its own container, connectable to BeLive Studio model providers.
Memoh is an open-source, self-hosted AI agent platform. Each bot runs in an isolated container with persistent memory and its own filesystem, supporting channels like Telegram, Discord, Feishu, QQ and Web UI.
- Official site: https://memoh.sh
- Docs: https://docs.memoh.ai
- Repository: https://github.com/memohai/Memoh
Prerequisites
- Memoh deployed (
curl -fsSL https://memoh.sh | sudo sh, Docker required) - Base URL:
https://api.tokenhub.belivestudios.com - API Key: create one in "API Tokens" on the BeLive Studio Console
Setup
- Open the Memoh web admin panel
- Go to "Model Providers" and add a custom provider of type OpenAI-compatible
- Fill in:
- Base URL:
https://api.tokenhub.belivestudios.com - API Key: your BeLive Studio API key
- Base URL:
- Add model IDs (e.g.
gpt-5.6-sol,claude-fable-5) - Select a model from this provider in the bot config and save
FAQ
Bot replies timing out?
Memoh runs inside a Docker network — make sure containers can reach api.tokenhub.belivestudios.com (test with curl inside a container).
Embedding models for memory/knowledge base?
Add an embedding model (e.g. text-embedding-3-small) to the provider and select it in the memory settings.