Course 1: MIFY Fundamentals — Chapter 3

Chapter 3: Adding AI to Your Workflow

Let’s connect to real AI services and understand provider options.

What You’ll Learn

• How to connect AI providers (OpenAI, Ollama, etc.)
• How to choose between Fast, Balanced, and Accurate modes
• How to generate images and text

AI Providers in MIFY

MIFY connects to 108+ AI services. The main options:

Provider	Cost	Setup	Best For
Ollama (local)	Free	Install Ollama, pull a model	Privacy, no API key needed
OpenAI	Pay per use	Add API key in Settings	GPT-4o, DALL-E
Anthropic	Pay per use	Add API key	Claude models
Google Gemini	Free tier available	Add API key	Gemini models
Cloudflare Workers AI	Free tier	Cloudflare account	Edge AI

Setting Up a Provider

1. Go to Settings (gear icon in the top right)
2. Navigate to Credentials
3. Click Add Credential
4. Select your provider (e.g., OpenAI)
5. Enter your API key
6. Click Save

Your key is encrypted and stored securely. You can use it in any workflow.

Provider Tiers

When running a workflow, you can choose a quality tier:

Tier	Speed	Cost	Quality
Fast	Fastest	Lowest	Good for drafts and prototypes
Balanced	Medium	Medium	Good for most tasks
Accurate	Slowest	Highest	Best for production quality

Text Generation

Use the Chat node:

1. Drag a Chat node onto the canvas
2. Set the provider and model
3. Write a system prompt (instructions for the AI)
4. Connect to a trigger
5. Run — type your question and see the AI respond

Image Generation

Use the Content Generation page at /content-gen:

1. Browse available image generation workflows
2. Adjust parameters (style, size, quality)
3. Click Generate
4. View results in your history

Exercise

1. Set up Ollama (free, local): install from ollama.com, run ollama pull phi3:mini
2. Build a workflow: Manual Trigger → Chat (using Ollama / phi3:mini)
3. Run it and have a conversation
4. Try the same workflow with a different model
5. Go to /content-gen and try generating an image

Key Takeaways

• Multiple AI providers available — from free local to paid cloud
• API keys are encrypted and manageable in Settings
• Provider tiers let you trade off speed vs quality
• Content generation has its own dedicated page

Previous: Chapter 2 — Nodes and Connections | Next: Chapter 4 — Human Approval Gates