--- title: "AI Image Generator Integration" description: "Generate images from text prompts with OpenAI DALL-E. FlowRunner agents produce visuals on demand and pass the hosted URL straight into downstream flow steps." url: https://flowrunner.ai/integrations/ai-image-generator date_modified: 2026-09-04T14:00:23-07:00 --- # AI Image Generator [AI](https://flowrunner.ai/integrations/category/ai-llms) Generate images from text prompts with OpenAI DALL-E. FlowRunner agents produce visuals on demand and pass the hosted URL straight into downstream flow steps. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 1 action · API key · available [AI Image Generator website](https://openai.com/) · [Platform Documentation](https://platform.openai.com/docs/guides/images) · Capability data verified 2026-07-21 1. A new row lands in a content sheet with a prompt column 2. Agent reads the prompt, target size, and quality from the row 3. Agent checks whether the row is flagged brand-sensitive before anything publishes 4. Image generated via DALL-E and returned as a hosted URL 5. URL written back to the source row for the content pipeline 6. Design channel notified that a new asset is ready 7. Brand-sensitive prompts routed to a reviewer before the image publishes ## What This Integration Enables Agents generate images from natural-language prompts without leaving the workflow. The connector returns a hosted URL, so the output feeds directly into a transactional email, a spreadsheet cell, a chat message, or a storage upload. It is a one-action surface on purpose: there is no library, no revision history, and no lookup that tells a flow what the model will accept, so size, quality, and model settings are declared on the call itself and the result is judged after it exists. The connector covers on-demand asset creation, not asset management; the surrounding flow decides where each image goes. ### Without FlowRunner **Bottlenecked on design**: Every routine visual waits in a designer's queue behind higher-priority work **Disconnected assets**: Generated images live in a chat thread and get copied around by hand **No pipeline**: Prompt, image, and the system that uses it are three separate manual steps ### With FlowRunner **Visuals on demand**: Routine images are produced inline the moment the flow needs them **URL feeds the next step**: The hosted image URL passes directly into email, storage, or chat **Reviewer sees brand-sensitive work**: A human approves the cases that carry brand risk before they go live ## Use Case Scenarios ### Content Pipeline Image Generation A content team keeps a planning sheet where each row holds a headline and an image prompt. A Google Sheets new-row trigger fires. The agent reads the prompt along with the size and quality declared on the row and calls Generate Image. It writes the returned URL back into the row's image column and posts a note to the design Slack channel that the asset is ready. The marketing operations lead moves from waiting on a queue to reviewing finished candidates. ### Personalized Email Visuals A customer milestone fires in the e-commerce system. The agent builds a prompt from the customer's product and context, calls Generate Image to produce a personalized visual, and passes the hosted URL to Brevo's Send Transactional Email so the image renders inline. The notification carries a relevant graphic without a designer touching each individual send. ### Design Review Gate for Brand Work A campaign prompt is flagged as brand-sensitive in the request. The agent generates a draft image but does not publish it. It posts the image URL and the originating prompt to the brand reviewer with a single question. The reviewer approves, requests a regeneration, or rejects. Only approved visuals advance to the publishing step. ## Human-in-Loop Highlight DALL-E output is an input to a human decision, not a substitute for one. When a prompt is tagged brand-sensitive, the agent generates the candidate image and then pauses, routing it to the brand reviewer via Slack: "Generated image for the spring campaign hero. Prompt: '\[prompt text\]'. Approve for publish, regenerate, or reject?" The reviewer sees exactly what the model produced and decides what happens next. Routine assets flow through automatically; the visuals that carry brand risk get a human in the loop before they ship. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via Slack Human decides Agent resumes with decision ## Agent Capabilities 1 actions ### Image Generation 1 - **Generate Image** Generates an image from a text prompt using OpenAI DALL-E or GPT-Image models and returns a hosted file URL, with size, quality, and model-specific settings supplied on the call. Used whenever a flow needs a visual on demand, from product mockups to email graphics, with the URL passed straight to the next step. ## Frequently Asked Questions ### What can FlowRunner do with AI Image Generator? FlowRunner agents can run Generate Image in AI Image Generator. ### Does connecting AI Image Generator to FlowRunner require OAuth? No. AI Image Generator connects to FlowRunner with an API key, no OAuth flow required. ### Can AI Image Generator trigger a FlowRunner workflow automatically? AI Image Generator doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at AI Image Generator?** This integration exposes AI Image Generator to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/ai-image-generator. Site index: https://flowrunner.ai/llms.txt