Answe

Feature

AI Image Backgrounds — Generate Widget Backgrounds From a Prompt

Type a prompt, pick a model, get a background. OpenAI and Google Gemini built into the Answe builder — you bring the key, we charge zero markup.

Free plan. No credit card required.

What AI Image Backgrounds gives you

Three models, one panel

Nano Banana 2, Nano Banana and ChatGPT — switch in a click.

Bring your own key

Pay your AI provider directly. No platform markup on generation.

Iterate cheap, finish big

512 → 1K → 2K → 4K size ladder so you don’t pay for drafts.

Made for widgets

Default 9:16 portrait fits the question widget out of the box.

How it works

AI Image Backgrounds — Generate Question Backgrounds From a Prompt

Introduction

Generate the perfect background image for any question without leaving the Answe builder. Type a prompt — "soft pastel sunrise over a coffee shop terrace", "modern IT consulting office, daylight" — pick an AI model, and click Generate. Preview the result, click Use as background, and your widget is ready. No stock-photo hunting, no design tools, no waiting on a designer.

You bring your own API key from OpenAI or Google Gemini, so generation costs go straight to your provider account at their listed rate — Answe never marks them up or charges for image generation itself.

How It Works

Three models to choose from

Each model has a different sweet spot. Switch any time without leaving the panel.

  • Nano Banana 2 — Google's gemini-3-pro-image-preview. Highest quality and the most controllable. Good for the final hero image.
  • Nano Banana — Google's gemini-2.5-flash-image. Faster and cheaper. Good for quick exploration.
  • ChatGPT — OpenAI's gpt-image-1. Strong on photographic and product-style imagery.

Pick the right size and shape

Two pickers right under the model row:

  • Aspect: 9:16 (default, portrait — matches most question widgets), 1:1 (square), 16:9 (landscape).
  • Size (Nano Banana 2 only): 512, 1K, 2K, 4K. Start at 512 while you iterate on prompt wording, then bump to 1K or higher when you've got a winner. The cost scales with size, so the iteration ladder saves money.

The Nano Banana model is fixed at 1024px and ChatGPT picks size automatically from your aspect choice — both hide the size picker so you don't have to think about it.

Bring your own keys

API keys live on your Answe account, in Settings → AI Image Generation Keys. Paste your OpenAI key, your Gemini key, or both. We store them on your account and never share them. Once a key is saved, the matching model in the builder works for any question you own.

If a key is missing or rejected, Answe shows a clear inline message with a link back to Settings — no cryptic errors.

Every generation is saved

The image lands on Answe's CDN (img1.qst.io) and a record is written to your media library with the original prompt, model, aspect ratio and size. That means:

  • The URL is permanent — no expiry, no stale links.
  • You always know exactly what produced an image.
  • Future updates can resurface generations as a personal gallery without losing history.

Apply with one click

When the image looks right, click Use as background. The widget canvas updates instantly and the background URL is committed to the question. Save the widget and it's live for visitors.

Key Benefits

  • Zero markup on AI cost — You bring your own key; Answe takes nothing on generation.
  • Iterate cheap, finish big512 → 1K → 2K → 4K ladder lets you explore prompts at low cost and only spend on the final render.
  • Three models, one panel — Switch between Google's two image models and OpenAI's gpt-image-1 without leaving the builder.
  • Made for widgets — Default 9:16 portrait fits the question widget out of the box; 1:1 and 16:9 for hero or landscape layouts.
  • Friendly errors — Rate limits, quota exhaustion and content-policy blocks come back with the actual provider message and a clear next step.
  • Provenance saved — Every generation is recorded in your media library with the prompt, model and settings used.
  • Secure key handling — Keys are stored on your account, never echoed back to the browser; the UI only ever shows the last four characters.

Getting Started

  1. Go to Settings (/account) and paste an OpenAI or Google Gemini API key under AI Image Generation Keys.
  2. Open any question in the builder and click the Image tab in the side panel.
  3. Pick a model: Nano Banana 2, Nano Banana or ChatGPT.
  4. Pick an aspect (default 9:16) and a size (Nano Banana 2 only — start with 512 for quick tests).
  5. Type a prompt and click Generate.
  6. Preview the result. Tweak the prompt and re-generate, or click Use as background to apply it.
  7. Save the widget — the background is live on every embed.

Tips

  • Start at 512 when you're still searching for a look. Once the prompt produces something close to right, bump to 1K.
  • Lean on Nano Banana 2 for hero shots, Nano Banana for quick rough drafts.
  • Use 9:16 for portrait widgets, 16:9 for landing-page heroes, 1:1 for tile-style embeds.
  • Be specific — "warm Helsinki café terrace at golden hour, soft light, shallow depth of field" beats "café".
  • Hit a quota error? Check your provider plan or wait — Answe shows the provider's actual rate-limit message so you know exactly what to do.

Related

  • Widgets — How backgrounds are applied to each widget type.
  • Template Gallery — Pre-built starting points if you'd rather not start from a blank canvas.
  • AI Agents — Combine AI-generated backgrounds with AI-powered conversation flows.
  • Analytics & Insights — See which question backgrounds drive the best engagement.

Included on every plan

Free€0100 responses / mo
Premium€19.9910k responses / mo
Pro€149.99100k responses / mo
EnterpriseCustomUnlimited
See all plans

Frequently asked questions

How much does it cost?

You pay your AI provider directly at their listed rate. Answe takes zero markup on image generation.

Where do I add my API keys?

In Answe Settings → AI Image Generation Keys. Paste your OpenAI and/or Google Gemini key once.

Are my keys safe?

Keys live on your account and are never echoed back to the browser. The UI only shows the last four characters.

Can I re-use a generated image later?

Yes. Every generation is recorded in your media library with the original prompt, model and settings.

Add AI Image Backgrounds to your site

Free plan, no credit card. Paste one embed snippet and you\u2019re live.

Start free