AI Model Providers

Configure local and cloud AI providers for Jarvis

Local-First Philosophy

Jarvis prioritizes local AI with Ollama — no API keys needed, no internet required, and your data stays private. Cloud providers are optional and can be added later.

Local AI with Ollama (Recommended)

Run AI models directly on your machine with Ollama:

Install Ollama:

curl -fsSL https://ollama.com/install.sh | sh

Download a model:

ollama pull llama3.2

Test it:

jarvis agent --message "Hello!" --local

Popular Ollama Models

Model	Size	Best For
llama3.2	4GB	General purpose, fast
qwen2.5	7GB	Coding, technical tasks
mistral	4GB	Balanced performance
codellama	7GB	Code generation

Cloud Providers (Optional)

Add cloud providers for more powerful models or when you need specific capabilities:

OpenAI

Set your API key:

export OPENAI_API_KEY="sk-..."

Or add to config:

jarvis configure

Anthropic (Claude)

Set your API key:

export ANTHROPIC_API_KEY="sk-ant-..."

Other Providers

Jarvis supports additional providers through the onboarding wizard:

OpenRouter - Access to multiple models through one API
AI Gateway - Vercel AI Gateway support
Moonshot - Chinese AI provider
Gemini - Google's AI models
Custom endpoints - Any OpenAI-compatible API

Configuration

Configure providers through the onboarding wizard or manually:

Run onboarding to add providers:

jarvis onboard

Or configure manually:

jarvis configure

Environment Variables

Variable	Description
OLLAMA_HOST	Ollama server address (default: 127.0.0.1:11434)
OPENAI_API_KEY	OpenAI API key
ANTHROPIC_API_KEY	Anthropic (Claude) API key

Model Configuration →

Learn about model selection and configuration

Configuration Guide →

Complete configuration reference