providers.md

January 30, 2026 · View on GitHub

Setup Guides

Setup: Ollama (Local, Privacy-First)

Recommended for: Healthcare, legal, finance, privacy-conscious users

1. Install Ollama

macOS:

brew install ollama

Linux:

curl https://ollama.ai/install.sh | sh

Windows: Download from https://ollama.ai/download

2. Start Ollama Service

ollama serve

3. Download a Model

Recommended models:

# Mistral (7B) - Best balance of speed/quality
ollama pull mistral:latest

# Llama 3 (8B) - Good for general tasks
ollama pull llama3:latest

# CodeLlama (7B) - Better for technical scheduling
ollama pull codellama:latest

4. Configure Nylas CLI

nylas ai config set default_provider ollama
nylas ai config set ollama.model mistral:latest
nylas ai config set ollama.host http://localhost:11434

5. Test

nylas calendar ai schedule "30-min call tomorrow afternoon"

Your data never leaves your machine!

Remote Ollama Hosts

If you're running Ollama on a different machine (e.g., a home server), configure the host:

# Configure remote Ollama host
nylas ai config set ollama.host http://192.168.1.100:11434

# Or using hostname
nylas ai config set ollama.host http://ollama-server.local:11434

# Test connection
curl http://192.168.1.100:11434/api/tags

Benefits of Remote Ollama:

  • Run on more powerful hardware (GPU-enabled server)
  • Share one Ollama instance across multiple machines
  • Offload AI processing from laptop

Security Note: Ensure your Ollama host is on a trusted network (LAN) and not exposed to the internet.


Setup: Claude (Cloud, Advanced)

Recommended for: Complex reasoning, long context analysis, multi-step workflows

1. Get API Key

Visit https://console.anthropic.com/ and create an API key.

2. Set Environment Variable

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

Add to ~/.bashrc or ~/.zshrc:

echo 'export ANTHROPIC_API_KEY="sk-ant-..."' >> ~/.bashrc
source ~/.bashrc

3. Configure Nylas CLI

nylas ai config set default_provider claude
nylas ai config set claude.model claude-3-5-sonnet-20241022

4. Test

nylas calendar ai schedule "Find time for team planning next week"

⚠️ Note: Data is sent to Anthropic's API (subject to their privacy policy)


Setup: OpenAI (Cloud, General)

Recommended for: General-purpose tasks, wide model selection

1. Get API Key

Visit https://platform.openai.com/api-keys

2. Set Environment Variable

export OPENAI_API_KEY="sk-..."
echo 'export OPENAI_API_KEY="sk-..."' >> ~/.bashrc
source ~/.bashrc

3. Configure Nylas CLI

nylas ai config set default_provider openai
nylas ai config set openai.model gpt-4-turbo

4. Test

nylas calendar ai schedule "Schedule meeting with john@example.com"

Setup: Groq (Cloud, Fast)

Recommended for: Real-time applications, low-latency requirements

1. Get API Key

Visit https://console.groq.com/

2. Set Environment Variable

export GROQ_API_KEY="gsk_..."
echo 'export GROQ_API_KEY="gsk_..."' >> ~/.bashrc
source ~/.bashrc

3. Configure Nylas CLI

nylas ai config set default_provider groq
nylas ai config set groq.model mixtral-8x7b-32768

4. Test

nylas calendar ai schedule "Quick 15-min sync tomorrow"