AI Image Generator (no API key)

Generate images from plain text prompts. No API key, no account, no card. Give it a list of prompts, get back a dataset with a permanent CDN image URL for each one.

This Actor is a thin wrapper over the anonymous free tier of Veida — the same free AI image generator the site serves in the browser, reachable here without a login step.

Why this one

Most image Actors and automation templates ask for a paid OpenAI, fal or Replicate key before they run once. This one runs on an empty account: it mints a throwaway anonymous client id per prompt and the free tier is metered against that.

Input

Field	Type	Notes
prompts	array of strings	One image per prompt. Required.
aspectRatio	enum	1:1 · 16:9 · 9:16 · 4:3 · 3:4. Default 1:1.
timeoutSecs	integer	Give up on one image and move to the next prompt. Default 240.

{
  "prompts": [
    "a paper boat on a calm lake at sunrise, soft light",
    "an isometric illustration of a small coffee shop, pastel palette"
  ],
  "aspectRatio": "16:9"
}

Not sure what to write? The AI image prompt library has ready-made prompts you can paste straight into the prompts list.

Output

One dataset row per prompt:

{
  "prompt": "a paper boat on a calm lake at sunrise, soft light",
  "aspectRatio": "16:9",
  "imageUrl": "https://cdn.veida.ai/uploads/kie/image/....webp",
  "watermarked": true
}

A prompt that fails still writes a row, with imageUrl: null and an error string — a silently shorter dataset is worse than a row that says why.

How it works

1. Posts the prompt to the generation endpoint that powers Veida's ChatGPT image generator, with a fresh anonymous client id.
2. Polls until the job reports success. Measured end to end: 40–120 seconds per image.
3. Writes the hosted image URL to the dataset. The URL is permanent and cacheable, so you can pass it straight to Google Sheets, Airtable, a CMS or a Slack message.

Limits of the free tier

Output is 1K and watermarked, and the anonymous allowance is capped per client and per IP per day — a long prompt list will eventually hit the wall, and the Actor reports that rather than failing silently. See what the free tier includes.

Signing in removes the watermark, raises the allowance and unlocks the heavier models — GPT Image 2 for text rendering, and Nano Banana 2 Lite image editor when you want to edit an existing image instead of generating from a prompt alone. There are example images if you want to judge output quality before running anything.

Notes

• Nothing is stored on our side beyond the generated image: the anonymous client id is created per prompt and thrown away.
• The endpoints are public and unauthenticated; no credentials are read from your Apify account.