Your First Build in 5 Minutes

From zero to working code. No prior setup needed. Follow these 5 steps.

Create Your Free Account

30 seconds

Create Account →

💡 Use a work email for team features later

💡 Your API key is generated automatically

Get a Free AI API Key

2 minutes

CesaFlow uses your own AI model key (BYOM). Groq, Cerebras, and Google Gemini all offer free tiers.

# Recommended free options:
# 1. Groq    → console.groq.com     (fastest, free tier)
# 2. Cerebras → cloud.cerebras.ai    (1M tokens/day free)
# 3. Gemini  → aistudio.google.com  (250 req/day free)
# 4. OpenRouter → openrouter.ai     (24+ free models)

Browse Free Models →

💡 Groq is the fastest free option

💡 You can switch models anytime in Dashboard → Model Settings

Configure Your Model

30 seconds

Go to Dashboard → Model Settings → Click your provider → Paste your API key → Save.

Model Settings →

💡 Set your preferred provider as "Default"

💡 CesaFlow encrypts your key with AES-256

Start Your First Run

1 minute to start, 5-15 minutes to complete

Go to Dashboard → Execute → Describe what you want to build → Click "Execute". Watch 4+ agents work in parallel.

# Example objectives to try:

"Build a REST API for a todo app with authentication"

"Create a landing page for a SaaS product with pricing"

"Build a CLI tool that converts CSV to JSON"

# Or use Goal Mode for complex projects:
"Build a complete project management SaaS with
 teams, tasks, Stripe billing, and deploy config"

Start Building →

💡 Start simple for your first run

💡 Use Goal Mode for multi-step projects

💡 Watch the Thinking tab to see agents reason in real-time

Review & Download

2 minutes

When the run completes, review generated files, check test results, and download your workspace as a ZIP.

# After your run completes:
# 1. Review files in the Files tab
# 2. Check test results in the Timeline
# 3. Download workspace as ZIP
# 4. Or open in CesaFlow IDE for editing

# Deploy with one command:
npx vercel deploy  # or railway up, fly deploy, docker compose up

View Runs →

💡 Use the Code Review feature to get AI feedback on the generated code

💡 Deploy configs are auto-generated when you use "with_devops" mode