B
AiderRun a Viber bot with zero data leaving your server
Private Viber bot on a self-hosted local LLM
setuproll@setuproll79.0Overall score
A Viber bot served entirely by a local Llama model through Ollama, so no message ever touches a third-party API. Aimed at privacy-sensitive groups or regions where Viber dominates and sending chat data to a US cloud is a non-starter.
79.0Score
640Votes
5Components
Install this build
terminal
ollama run llama3.3:70bComponents
Model
- Llama 3.3 70B (Ollama)
Stack
- Ollama
- FastAPI
- Viber REST bot API
Tools
- Viber webhook
- Caddy reverse proxy
- systemd service
Hosting
- Single GPU VPS
- no external API
- fully self-hosted
How it works
- Viber posts incoming messages to the FastAPI webhook
- FastAPI forwards the prompt to the local Ollama server
- Llama 3.3 70B generates the reply on your own hardware
- Response is sent back through the Viber send-message endpoint
Summary
A Viber bot served entirely by a local Llama model through Ollama, so no message ever touches a third-party API. Aimed at privacy-sensitive groups or regions where Viber dominates and sending chat data to a US cloud is a non-starter.
79.0 score 640 votes
0 Reviews
Your rating
Sign in to post
Loading discussion...