KIE.AI + Google Sheets: consistent character images
Your mascot looks perfect in one image, then completely different in the next. New face shape, weird colors, “close enough” outfits. And suddenly you’re stuck rerunning prompts, hunting down the right version, and explaining to a teammate which file is the “real” one.
This is where KIE.AI image automation helps. Marketing managers feel it when campaigns need fast approvals. Designers feel it when variations multiply. Founders feel it when brand consistency turns into late-night rework.
This workflow lets you request character-consistent images from KIE.AI (Nano Banana API) through a simple form, then capture the output in Google Sheets so you can review, reuse, and scale without chaos.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: KIE.AI + Google Sheets: consistent character images
flowchart LR
subgraph sg0["Submit Text Prompt for image Generation Flow"]
direction LR
n0["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/>Obtain the generated status"]
n1["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/form.svg' width='40' height='40' /></div><br/>Submit Text Prompt for image.."]
n2["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/>Send image Generation Reques.."]
n3@{ icon: "mdi:cog", form: "rounded", label: "Wait for image Processing Co..", pos: "b", h: 48 }
n4@{ icon: "mdi:swap-horizontal", form: "rounded", label: "Check if image Generation is..", pos: "b", h: 48 }
n5@{ icon: "mdi:swap-vertical", form: "rounded", label: "Format and Display image Res..", pos: "b", h: 48 }
n0 --> n4
n3 --> n0
n4 --> n5
n4 --> n3
n1 --> n2
n2 --> n3
end
%% Styling
classDef trigger fill:#e8f5e9,stroke:#388e3c,stroke-width:2px
classDef ai fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
classDef aiModel fill:#e8eaf6,stroke:#3f51b5,stroke-width:2px
classDef decision fill:#fff8e1,stroke:#f9a825,stroke-width:2px
classDef database fill:#fce4ec,stroke:#c2185b,stroke-width:2px
classDef api fill:#fff3e0,stroke:#e65100,stroke-width:2px
classDef code fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
classDef disabled stroke-dasharray: 5 5,opacity: 0.5
class n1 trigger
class n4 decision
class n0,n2 api
classDef customIcon fill:none,stroke:none
class n0,n1,n2 customIcon
The Problem: Character “Consistency” Is Usually Manual
Character consistency breaks in boring, expensive ways. You tweak a prompt to get a different pose, and the face changes. You try a new background, and suddenly the outfit design drifts. Then the review cycle starts: someone downloads three versions, drops them into Slack, and asks “which one matches the last post?” Multiply that by a week of content, and you’re doing more asset wrangling than creative work. The worst part is the mental load. You’re constantly second-guessing if the next generation will match your “canon” character.
The friction compounds. Here’s where it usually falls apart.
- Prompt reruns pile up because you don’t have a repeatable request process for the same character.
- Reviews get slower when outputs live in random downloads folders or chat threads.
- Image-to-image attempts become guesswork if reference URLs aren’t tracked alongside the prompt.
- Costs sneak up because “one more generation” becomes the default fix for inconsistency.
The Solution: A Simple Form That Generates and Logs Consistent Images
This n8n workflow turns character image requests into something you can run like a process, not a gamble. You submit a form with the model you want (text-to-image or image-to-image), a prompt, and optionally a set of reference image URLs. n8n sends that request to KIE.AI’s Nano Banana API, waits briefly, then polls for completion every few seconds until the generation is ready. Once it’s done, the workflow formats the result into clean fields you can store and review. In practice, you stop losing time to “where did that version go?” and you start building a reliable library of on-brand character visuals.
The workflow starts with a form submission. KIE.AI handles the generation in the background while n8n waits and checks status. Finally, the output is packaged in a tidy format that’s ready to log in Google Sheets and share with your team.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say your team needs 15 character variations a week for ads, landing pages, and social. Manually, it’s easy to spend about 10 minutes per variation just copying prompts, tracking which reference image you used, and posting results for review (around 2.5 hours weekly). With this workflow, submitting the form takes a minute, and KIE.AI typically finishes in about 10–20 seconds while n8n polls automatically. You still review the output, but the “admin work” mostly disappears, so the same 15 requests feel manageable instead of draining.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- KIE.AI (Nano Banana API) for character-consistent image generation
- Google Sheets to log prompts and outputs
- KIE.AI API key (get it from your KIE.AI Nano Banana dashboard)
Skill level: Beginner. You’ll paste an API key, connect Google Sheets, and edit a couple of form fields.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
Image request comes in through a form. You choose the model (text-to-image or edit mode) and submit a prompt. If you’re using image-to-image, you can also add up to five reference image URLs.
The workflow sends the generation request to KIE.AI. n8n uses an HTTP Request step to call the Nano Banana API with your settings and API key, which means you’re not manually jumping between tools.
n8n waits, then checks status automatically. A short Wait step pauses the workflow, then another request asks KIE.AI if the job is finished. If it’s still processing, the workflow loops and checks again.
Results are prepared for logging and review. Once the generation is complete, the output is cleaned up in a “set fields” step so it can be written into Google Sheets and shared internally without extra formatting work.
You can easily modify the form fields to capture extra metadata (campaign name, channel, due date) based on your needs. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Form Trigger
This workflow starts with a user-submitted form that collects the image prompt details and API key.
- Add the Image Prompt Form Trigger node and open its settings.
- Set Form Title to
AI image generator. - Set Form Description to
Please fill in the following information to generate your video. - Ensure the form fields include prompt, img_url, model, and api_key.
Step 2: Connect the Image Generation API Call
This step sends the form data to the image generation API to create a task.
- Add the Dispatch Image Generation Call node and connect it to Image Prompt Form Trigger.
- Set URL to
https://api.kie.ai/api/v1/playground/createTask. - Set Method to
POSTand enable Send Body and Send Headers. - Set Specify Body to
jsonand paste the JSON body as={ "model": "{{$json.model}}", "callBackUrl": "https://your-domain.com/api/callback", "input": { "prompt": "{{$json.prompt}}", "image_urls": [{{$json.img_url ? $json.img_url.split(',').map(url => '"' + url.trim() + '"').join(',') : ''}}] } }. - Add header parameters: Content-Type =
application/jsonand Authorization ==Bearer {{$json.api_key}}.
⚠️ Common Pitfall: Replace https://your-domain.com/api/callback with your real callback URL if your API account requires it.
Step 3: Set Up the Processing Loop
The workflow waits, checks status, and loops until the image generation succeeds.
- Add the Pause for Image Processing node and connect it to Dispatch Image Generation Call.
- Add the Retrieve Generation Status node and connect it to Pause for Image Processing.
- In Retrieve Generation Status, set URL to
https://api.kie.ai/api/v1/playground/recordInfo. - Enable Send Query and add a query parameter taskId with value
={{$json.data.taskId}}. - Add header parameters: Content-Type =
application/jsonand Authorization ==Bearer {{$node["Image Prompt Form Trigger"].json["api_key"]}}. - Add the Verify Generation Completion node and connect it to Retrieve Generation Status.
- In Verify Generation Completion, set the condition to compare Left Value
=truewith Right Value={{ $json.data.state == 'success'}}. - Ensure Verify Generation Completion routes success to Prepare Image Result Output and failure back to Pause for Image Processing.
Step 4: Configure the Output Node
This step formats the successful image result for downstream use.
- Add the Prepare Image Result Output node and connect it to the success output of Verify Generation Completion.
- In Prepare Image Result Output, add an assignment with Name
imgand Value={{ $json.data.resultJson }}.
Step 5: Test and Activate Your Workflow
Validate the full loop and confirm the image output is returned correctly.
- Click Execute Workflow and submit the Image Prompt Form Trigger with a valid api_key.
- Confirm Dispatch Image Generation Call returns a task ID and Retrieve Generation Status polls the task.
- Verify the success path passes through Verify Generation Completion and outputs img in Prepare Image Result Output.
- When successful, toggle the workflow to Active to enable production use.
Common Gotchas
- KIE.AI credentials can expire or be pasted incorrectly. If things break, check the API key field in your form submission and the Authorization header in the HTTP Request node first.
- If you’re using Wait nodes or external rendering, processing times vary. Bump up the wait duration if downstream nodes fail on empty responses.
- Default prompts in AI nodes are generic. Add your brand voice early or you’ll be editing outputs forever.
Frequently Asked Questions
About 20 minutes if you already have your KIE.AI key and a Google Sheet ready.
No. You will connect accounts and paste an API key. The rest is editing form fields and picking where the rows go in Google Sheets.
Yes. n8n has a free self-hosted option and a free trial on n8n Cloud. Cloud plans start at $20/month for higher volume. You’ll also need to factor in KIE.AI API costs: 20 free generations, then about $0.02 per request.
Two options: n8n Cloud (managed, easiest setup) or self-hosting on a VPS. For self-hosting, Hostinger VPS is affordable and handles n8n well. Self-hosting gives you unlimited executions but requires basic server management.
Yes, and it’s honestly one of the best tweaks. Add a few extra fields in the form trigger (like campaign name and “needs review”), then extend the final “Prepare Image Result Output” step to include those columns before writing to Google Sheets. Many teams also add a “status” field (Draft, Approved, Needs changes) and a link to the reference image URLs used. That way your approvals stay attached to the exact generation recipe, not somebody’s memory.
Most of the time it’s an invalid or inactive API key, so regenerate it in KIE.AI and update the value used in the HTTP Request. Also check the selected model string (for example, the text-to-image vs edit model), because a mismatch can cause failed requests. If you submit multiple image URLs, make sure they’re public and properly separated, since broken links can lead to generation errors. Finally, if you’re doing a lot of runs in a short window, you might hit rate limits, so spacing executions out can help.
A lot. On n8n Cloud, your limit is mostly your monthly executions. If you self-host, there’s no hard execution cap, and the practical limit becomes your server resources and how fast KIE.AI returns results.
Often, yes, because this flow depends on polling and conditional looping, which is awkward (and sometimes expensive) in simpler automation tools. n8n handles “wait, check, repeat until complete” cleanly, so you don’t have to bolt together workarounds. You also get more control over how prompts, reference URLs, and outputs are normalized before they hit Google Sheets. Zapier or Make can still work if you only need a single API call and don’t care about status checks, but you’ll usually feel the limits quickly once you scale. Talk to an automation expert if you’re not sure which fits.
Once you can generate consistent character images on demand and log every output, the review process gets calmer. Set it up once, then let the workflow do the repetitive parts while you focus on the creative decisions.
Need Help Setting This Up?
Our automation experts can build and customize this workflow for your specific needs. Free 15-minute consultation—no commitment required.
