Image from Description
Describe the picture you need in plain words. The workflow writes a detailed prompt, then generates the image.
Workflow on the canvas
Nodes
- Image Description (Text Input)
- Enhance Prompt (AI Text)
- Generate Image (AI Image)
- Generated Image (Output)
How the nodes connect
- Image Description → Enhance Prompt
- Enhance Prompt → Generate Image
- Generate Image → Generated Image
What it does
Make visuals for posts, articles, and campaign drafts without writing prompts yourself. You describe the idea in a sentence; the first AI node expands it into a detailed prompt covering style, lighting, composition, and mood, and the AI Image node generates a square image from it. The prompt-writing node leaves out your project context by default, so copy rules do not crowd the image prompt. Each node's output is recorded in History, so you can see which prompt produced an image.
Who it is for
Content marketers, bloggers, and social media managers who need illustrations or concept images and would rather not learn prompt writing.
What you need
A short description of the image, typed into the Image Description node: for example the subject, the setting, and the feeling you want.
Steps on the canvas
- Image Description (Text Input): describe the image in a sentence or two.
- Enhance Prompt (AI Text): turns your description into a detailed image generation prompt covering style, lighting, composition, and mood.
- Generate Image (AI Image): generates a 1024×1024 image from the prompt, in the Natural style by default. Change the size or style in the Inspector.
- Generated Image (Output): shows the generated image.
Running it in batches
To make one image per idea, replace the Text Input node with a Dataset node, choose an entry set with one description per row (for example, a column named description), save the workflow, and click Run batch. The workflow runs once per row.
Example
Example input
A quiet coffee shop on a rainy morning, seen from a window seat.
Example output
Prompt written by Enhance Prompt: "Cozy coffee shop interior on a rainy morning, view from a window seat, raindrops on the glass, soft diffused daylight, warm wooden tables, steam rising from a ceramic cup, muted earthy palette, shallow depth of field, calm and reflective mood." Generate Image then turns this prompt into a square image.
Illustrative example. Actual output depends on the model and your instructions.
Start with this template
Create an account to open this template on the canvas. During the preview, an invite code is required to use the product.
Editing on the canvas needs a computer or a wide screen.
More templates
- Simple Text Enhancement — Enhance or rewrite text using AI
- Batch Content Generator — Generate content for multiple entries from a dataset
- Context-Aware Generation — Generate content using project context and guidelines
- Product Description + Image — Generate product descriptions and images from data
- Game Character + Scene Art — Turn world and character notes into a character illustration and a matching scene
- Brand Style Series — Three images in one shared style, with optional brand reference images
- Social Ad Variants — One brief, a different post for each platform, plus a square and a vertical image
- Blog SEO Article + Cover — Outline, write, and illustrate an SEO blog post from a topic and keywords
- Short-Form Script + Storyboard — A shot-by-shot script and three vertical key frames for TikTok, Reels, or Shorts