Sign inSign up

kapilyadav22/runlocalllm

By kapilyadav22

•Updated 20 days ago

100% Offline LLM UI: Arena, DeepSeek-R1, Vision, Voice, RAG, Sandbox, Live Benchmarks & Ollama

Image
API management
Machine learning & AI
Developer tools
0

126

kapilyadav22/runlocalllm repository overview

⁠LocalLLMMind 🧠

⁠Modern, Privacy-First Desktop AI Workstation for Local LLMs

100% Local • Zero Telemetry • Real-Time Streaming • Multi-Model Arena

Docker Pulls Docker Image Size Ollama License: Apache 2.0

⁠⚡ 10-Second Quick Start

Run LocalLLMMind and connect instantly to Ollama⁠ running on your machine:

docker run -d \
  --name localllmmind \
  -p 3000:80 \
  --add-host=host.docker.internal:host-gateway \
  -e OLLAMA_URL=http://host.docker.internal:11434 \
  --restart unless-stopped \
  kapilyadav22/localllmmind:latest

👉 Open http://localhost:3000⁠ in your browser.

(Make sure Ollama allows local cross-origin connections: OLLAMA_ORIGINS="*" ollama serve)

⁠🐳 Docker Compose

Prefer Compose? Add this to your docker-compose.yml:

services:
  localllmmind:
    image: kapilyadav22/localllmmind:latest
    container_name: localllmmind-app
    ports:
      - "3000:80"
    environment:
      - PORT=80
      - OLLAMA_URL=http://host.docker.internal:11434
    extra_hosts:
      - "host.docker.internal:host-gateway"
    restart: unless-stopped

docker compose up -d

⁠🚀 Why Choose LocalLLMMind?

FeatureWhat It Gives You
🔒 100% Offline & PrivateZero telemetry, no cloud accounts, no API fees. All chats and embeddings stay on your hardware.
⚔️ Model Arena ModePrompt two local models side-by-side with synchronized input and live speed/token benchmarking.
⚡ Hardware AccelerationMicro-batched token rendering delivers buttery-smooth 30–80+ tok/s streaming.
🧠 Reasoning Model SupportNative chain-of-thought parsing with interactive <think> collapse for DeepSeek-R1 and Qwen.
👁️ Vision & MultimodalDrag-and-drop image analysis with llama3.2-vision, llava, and moondream.
📄 Chat with Documents (RAG)Ingest 40+ code, data, and document formats completely client-side.
✏️ Edit & Branch ChatsChatGPT-style message editing and multi-version response branching (◀ 1/3 ▶).
📥 Built-in Model ManagerSearch, pull, inspect, and delete Ollama models directly from the UI without terminal commands.
🧪 Live Artifact SandboxRender interactive HTML, React, SVG, and Mermaid diagrams directly in your chat stream.
🪶 Featherweight ImageMulti-stage Alpine Linux + Nginx image under 72 MB with instant startup.

⁠⚙️ Configuration

VariableDefaultDescription
PORT80Internal container port served by Nginx
OLLAMA_URLhttp://host.docker.internal:11434URL of your Ollama instance

⁠👨‍💻 Links & Author

Tag summary

Content type

Image

Digest

sha256:c3b003c73…

Size

10.6 MB

Last updated

20 days ago

docker pull kapilyadav22/runlocalllm