OpenChat-Cuda

OpenChat-Cuda

Docker app from el_pino's Repository

Overview

A self-hosted, offline, ChatGPT-like chatbot with open source LLM support. 100% private, with no data leaving your device. Please note that this version requires an NVIDIA GPU with the Unraid NVIDIA-DRIVER plugin.

Runtime arguments

Web UI
http://[IP]:[PORT:3000]/
Network
bridge
Shell
sh
Privileged
false
Extra Params
--gpus all

Template configuration

Local Model PathVariable

The local model path

Target
MODEL
Default
/models/llama-2-7b-chat.bin
Model Download URLVariable

GGML Model Binary.

Target
MODEL_DOWNLOAD_URL
Default
https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/resolve/main/llama-2-7b-chat.ggmlv3.q4_0.bin
Model DirectoryPathrw

The local model directory to use as a cache

Target
/models
Default
/mnt/user/appdata/models
Web UIPorttcp

Chat UI Port

Target
3000
Default
3000
API PortPorttcp

HTTP API Port

Target
8000
Default
8000
Number Of GPU LayersVariable

Layers to offload to GPU. Update this number if server fails to load.

Target
N_GPU_LAYERS
Value
12

Details

Repository
ghcr.io/edgar971/open-chat-cuda:v1.0.6
Last Updated2026-06-01
First Seen2023-09-06

Run OpenChat-Cuda on Unraid.

OpenChat-Cuda is listed in Community Apps for Unraid OS. Explore Unraid to build a flexible home server, NAS, or homelab.