Your First Workflow

What we'll build

A simple 2-node workflow: a TextInput feeds a prompt to Gemini, which generates a response.

TextInput ("Write a tagline for...") ──prompt──▶ Gemini ──response──▶ [artifact]

Step 1: Create the workflow

1. From your dashboard, click New Workflow
2. Give it a name (e.g., “My First Workflow”)
3. You're now in the visual editor — an empty canvas

Step 2: Add a TextInput node

1. Click the + button in the toolbar
2. Select Text Input from the Input category
3. Click the node to open its config panel
4. Type your prompt: Write a catchy tagline for a coffee brand called “MoonBrew”

Step 3: Add a Gemini node

1. Click + again
2. Select Gemini from the AI Text category
3. Position it to the right of TextInput
4. Default model is gemini-2.5-flash — leave it as is

Step 4: Connect them

1. Drag from the text output port of TextInput
2. Drop onto the prompt input port of Gemini
3. A cable appears — the connection is live

Step 5: Execute

1. Click the Execute button (top-right)
2. TextInput turns green instantly (it just outputs your text)
3. Gemini turns blue (processing) then green (done)
4. Click the Gemini node to see the generated tagline

What just happened

Behind the scenes:

1. Zephly validated your DAG (no cycles, all ports connected)
2. Checked your credit balance (Gemini costs 1-2 credits)
3. Executed TextInput → passed its output to Gemini's prompt
4. Called the Gemini API with your text
5. Stored the response as an artifact
6. Sent WebSocket events to update the UI in real-time

Next: make it more interesting

Try adding more nodes:

• Add a NanoBanana node → connect Gemini's response to its prompt → generate an image from the tagline
• Add an ElevenLabs node → generate a voiceover of the tagline
• Add an Instagram node → publish the image directly

Related

• Workflow Editor Guide
• Plugin Reference
• Understanding Credits