Skip to content

INSTALL.md

Latest commit

 

History

History
276 lines (189 loc) · 7.38 KB

INSTALL.md

File metadata and controls

276 lines (189 loc) · 7.38 KB

eVera — Installation Guide

eVera is fully self-contained. Every dependency — Python, Ollama, Playwright, system libraries, and Electron — is installed automatically by the setup scripts. No manual configuration is required to get started.

Quick Start

Linux / macOS — One Command

curl -fsSL https://raw.githubusercontent.com/embeddedos-org/eVera/main/install.sh | bash

This clones the repository, installs Python 3.12, creates a virtual environment, installs all Python packages, installs Ollama, pulls a default offline model, installs Playwright Chromium, and installs the Electron desktop dependencies.

Windows — One Command (PowerShell)

irm https://raw.githubusercontent.com/embeddedos-org/eVera/main/setup.ps1 | iex

After Setup — Start eVera

source .venv/bin/activate          # Linux/Mac
# .venv\Scripts\activate           # Windows

python main.py --mode server       # Start the web server
# Open http://localhost:8000

Download Pre-Built Installers

Pre-built installers are published automatically on every tagged release via GitHub Actions CI.

Platform File How to Install
Windows Vera-Setup-*.exe Run the installer — includes Python backend, no dependencies needed
macOS Vera-*.dmg Open DMG → drag Vera to Applications
Linux Vera-*.AppImage chmod +x Vera-*.AppImage && ./Vera-*.AppImage
Linux (deb) Vera-*.deb sudo dpkg -i Vera-*.deb
Android Vera-*.apk Enable "Install from unknown sources" → tap APK

Download from: github.com/embeddedos-org/eVera/releases

Operating Modes

eVera operates in three distinct modes, selectable from the header dropdown.

Mode Internet LLMs Available Computer Control LAN Access
LOCAL Not required Ollama, LM Studio, Jan, llama.cpp Full No
LAN Not required All offline + LAN-hosted servers Full Yes
WWW Required All 100+ models including cloud Full Yes

To set the default mode, edit .env:

VERA_SERVER_DEFAULT_MODE=local    # or lan, www

Offline LLMs

eVera works completely offline using Ollama. After setup, pull any model:

# Recommended starter models
ollama pull qwen3:8b          # 5 GB — best general-purpose
ollama pull llama3.2:3b       # 2 GB — fast, low RAM
ollama pull deepseek-r1:7b    # 5 GB — best reasoning
ollama pull qwen2.5-coder:7b  # 5 GB — best code

# Tiny models for very low RAM systems (< 4 GB)
ollama pull qwen3:0.6b        # 400 MB
ollama pull llama3.2:1b       # 700 MB
ollama pull moondream:1.8b    # 1.5 GB — vision

# Large models for powerful hardware
ollama pull qwen3:32b         # 20 GB
ollama pull llama3.3:70b      # 40 GB

LM Studio

  1. Download and install LM Studio
  2. Load any model in LM Studio
  3. Start the local server (default port 1234)
  4. In eVera, select any lmstudio/ model from the model picker

Configure the URL in .env if you use a non-default port:

VERA_LLM_LM_STUDIO_URL=http://localhost:1234

Jan AI

  1. Download and install Jan
  2. Load any model in Jan
  3. Enable the API server (default port 1337)
  4. In eVera, select any jan/ model from the model picker
VERA_LLM_JAN_URL=http://localhost:1337

llama.cpp Server

  1. Build or download llama.cpp
  2. Start the server: ./llama-server -m model.gguf --port 8080
  3. In eVera, select any llamacpp/ model from the model picker
VERA_LLM_LLAMACPP_URL=http://localhost:8080

Manual Installation (Step by Step)

If you prefer to install manually or the one-liner fails:

1. Clone the Repository

git clone https://github.com/embeddedos-org/eVera.git
cd eVera

2. Python Environment

python3.12 -m venv .venv
source .venv/bin/activate          # Linux/Mac
# .venv\Scripts\activate           # Windows

pip install --upgrade pip
pip install -r requirements.txt

3. System Packages (Linux)

sudo apt install -y \
  tesseract-ocr ffmpeg libportaudio2 \
  xdotool wmctrl scrot xclip

4. Playwright (Browser Automation)

python -m playwright install chromium --with-deps

5. Ollama (Offline LLMs)

curl -fsSL https://ollama.com/install.sh | sh
ollama pull qwen3:8b

6. Configuration

cp .env.example .env
# Edit .env to add optional cloud API keys
# All fields are optional — Ollama works without any keys

7. Start eVera

python main.py --mode server
# Open http://localhost:8000

Desktop App (Electron)

The Electron desktop app wraps the Python backend into a native window with system tray, global shortcut, and auto-start support.

Run from Source

# First, ensure the Python backend is running or let Electron start it
cd electron
npm install
npm start

Build Installer

# Linux
bash setup.sh --build-desktop
# Produces: electron/dist/Vera-*.AppImage and Vera-*.deb

# Windows (PowerShell)
.\scripts\build_windows.ps1
# Produces: electron\dist\Vera-Setup-*.exe

# macOS
cd electron && npm run build:mac
# Produces: electron/dist/Vera-*.dmg

LAN / Network Mode

To allow other computers on your network to access eVera:

python main.py --mode server --host 0.0.0.0

Then open http://<your-ip>:8000 from any device on the same network.

To require authentication for LAN connections:

VERA_SERVER_ZONE_LAN_AUTH_REQUIRED=true
VERA_SERVER_API_KEY=your-secret-key

Environment Variables Reference

All settings use the VERA_ prefix and are loaded from .env. No setting is mandatory — eVera runs fully offline with zero configuration.

Variable Default Description
VERA_LLM_OLLAMA_URL http://localhost:11434 Ollama server URL
VERA_LLM_LM_STUDIO_URL http://localhost:1234 LM Studio server URL
VERA_LLM_JAN_URL http://localhost:1337 Jan AI server URL
VERA_LLM_LLAMACPP_URL http://localhost:8080 llama.cpp server URL
VERA_LLM_OPENAI_API_KEY (empty) OpenAI API key (optional)
VERA_LLM_ANTHROPIC_API_KEY (empty) Anthropic API key (optional)
VERA_LLM_GEMINI_API_KEY (empty) Google Gemini API key (optional)
VERA_SERVER_HOST 0.0.0.0 Server bind host
VERA_SERVER_PORT 8000 Server port
VERA_SERVER_API_KEY (empty) API key for LAN/WWW auth

Troubleshooting

eVera starts but shows no models: Run ollama list to confirm models are installed. If empty, run ollama pull qwen3:8b.

LM Studio / Jan not detected: Ensure the local server is running and the port matches VERA_LLM_LM_STUDIO_URL / VERA_LLM_JAN_URL in .env.

Playwright install fails: Run python -m playwright install chromium --with-deps manually. Browser automation is optional — all other features work without it.

Port 8000 already in use: Change the port with python main.py --mode server --port 8001.

Windows: Python not found: Install Python 3.12 from python.org and check "Add Python to PATH" during installation.

Updating eVera

git pull origin main
pip install -r requirements.txt   # pick up new Python deps

Or re-run the one-liner installer — it updates an existing install automatically.