Bulk AI Image Generator — Generate Hundreds of AI Images from Prompts (OpenRouter)

Turn a list of text prompts into a gallery of AI-generated images in a single run. This Apify Actor lets you pick from the top image-generation models — Google Gemini 3.1 Flash / 3 Pro / 2.5 Flash Image ("Nano Banana", "Nano Banana Pro") and OpenAI GPT-5 / GPT-5.4 Image — then generates, stores, and tabulates every image for you. No API keys to manage: model usage runs on your Apify credit via the official apify/openrouter proxy, plus a small per-image fee.

Bulk image generation, batch AI art, prompt-to-image automation, text-to-image API, AI image generator at scale — all from one Actor.

---

✨ What it does

• Batch generation — feed it a list of prompts and it produces one (or many) images per prompt.
• Top models, your choice — select from a curated list of leading OpenRouter image models, or paste any custom model ID.
• Controllable output — set aspect ratio (1:1, 16:9, 9:16, …) and resolution (0.5K → 4K).
• Variations — generate up to 10 variations per prompt with one click.
• Clean results — every image is stored in the run's key-value store and exposed as a public URL in a tidy dataset table you can preview, export (CSV/JSON/Excel), or pull via API.
• Fast & concurrent — generates many images in parallel (configurable concurrency) instead of one at a time.
• Migration-safe — progress is persisted continuously, so if Apify migrates or restarts the run it resumes where it left off without re-generating (or re-charging for) finished images.
• Resilient — automatic retries with backoff on rate limits (429) and server errors; a single failed prompt is logged and skipped, never crashing the whole run.

🤔 Why use it

You already have prompts. You don't want to: write API glue code, handle base64 decoding, manage rate limits, or click "generate" 300 times. This Actor does the boring part and hands you a downloadable dataset of images with full metadata.

---

🚀 Quick start

1. Open this Actor on Apify. No API key needed — model access runs on your Apify credit through the apify/openrouter proxy.
2. Add your Prompts (one per line).
3. Pick a model, aspect ratio, and resolution.
4. Click Start. When the run finishes, open the Dataset tab to preview and download your images.

Input example

{
  "prompts": [
    "A photorealistic golden retriever puppy in a sunlit meadow, shallow depth of field",
    "A minimalist flat-design mountain logo inside a circle, blue and white",
    "A cyberpunk city street at night, neon reflections on wet pavement"
  ],
  "model": "google/gemini-2.5-flash-image",
  "systemPrompt": "Cohesive flat-design illustration style, soft pastel palette",
  "aspectRatio": "16:9",
  "imageSize": "2K",
  "imagesPerPrompt": 2,
  "maxItems": 50
}

Output sample

Each dataset item looks like this:

{
  "index": 0,
  "prompt": "A photorealistic golden retriever puppy in a sunlit meadow, shallow depth of field",
  "model": "google/gemini-3.1-flash-image-preview",
  "aspectRatio": "16:9",
  "imageSize": "2K",
  "imageUrl": "https://api.apify.com/v2/key-value-stores/XXXX/records/image-00000.png",
  "contentType": "image/png",
  "fileExtension": "png",
  "fileSizeBytes": 1742310,
  "description": "A golden retriever puppy sitting in tall grass...",
  "success": true
}

The imageUrl points at the image binary stored in the run's key-value store. The Overview dataset view renders these as thumbnails.

---

💰 Cost

You pay the Actor's flat per-image fee + the exact OpenRouter cost of each generated image, rounded up to the nearest $0.00001. The OpenRouter cost depends on the model and resolution you pick, so the total scales precisely with your usage.

• Actor fee — a small fixed charge per generated image (the apify-default-dataset-item pay-per-event, charged automatically for every image pushed to the dataset).
• OpenRouter cost — the real model/compute cost of the image, passed through and rounded up to the nearest $0.00001.

⚠️ Free Apify plan users are charged 10× the AI token (OpenRouter) cost. Upgrade to any paid Apify plan for standard (1×) token pricing. Free plans are also capped at 10 images per run.

---

🎯 Use cases

• E-commerce — generate product mockups, lifestyle shots, and ad creatives in bulk.
• Marketing & social — produce a month of on-brand social posts from a content calendar of prompts.
• Game & app assets — icons, textures, concept art at scale.
• Blogging & SEO — unique header images for every article.
• Design exploration — fan out a single idea into dozens of variations and pick the best.
• Datasets — build labeled synthetic image sets for ML experiments.

---

❓ FAQ

Do I need my own OpenRouter API key? No. By default, model access is routed through Apify's official apify/openrouter proxy and billed as Apify credit using your run token — there's nothing to sign up for or paste. Optionally, if you'd rather use your own OpenRouter account, paste your key into the OpenRouter API key field (under Advanced settings): the Actor will then call OpenRouter directly and bill your OpenRouter account instead of your Apify credit. You still pay the small Actor per-image fee.

Which models are supported? Any image-output model offered through the proxy. The dropdown curates the best ones (Gemini 2.5 Flash / 3.1 Flash / 3 Pro Image, GPT-5 / GPT-5.4 Image); the Custom model ID field accepts any other (see the image-model list).

Why did a prompt come back as text instead of an image? Some chat-capable image models will answer a terse prompt (e.g. "flying monkey") with a clarifying question instead of drawing. This Actor sends a built-in system instruction forcing the model to always render an image; you can reinforce it with your own System prompt. Gemini models are the most reliable for this — switch to one if an OpenAI model keeps replying with text.

Where are my images? Each image is saved in the run's key-value store and linked from the dataset via imageUrl. Export the dataset as JSON/CSV/Excel, or download images directly from the links.

What happens if a prompt fails or is rate-limited? The client retries automatically with backoff (respecting Retry-After). If it still fails, that single image is logged and skipped — the rest of the run continues and you are not charged for it.

Can I control spend? Yes — set Max images to hard-cap a run, and set a run-level spend limit in Apify. The Actor stops pushing once the budget is reached.

---

🔒 Responsible use

This Actor is a thin automation layer over OpenRouter's models. You are responsible for your prompts and the images you generate, and for complying with OpenRouter's terms and the underlying model providers' usage policies. Do not generate content that is illegal, infringing, deceptive, or that depicts real people without consent. Respect copyright and personality rights.

---

🛠️ Monetization setup (Actor owner)

After apify push, configure pricing in Apify Console → Actor → Publication → Monetization:

1. Pricing model: Pay per event.
2. Add events with names exactly matching the code:
   • apify-actor-start → keep the default price.
   • apify-default-dataset-item → the per-image fee (this is the synthetic per-result event; the platform charges it automatically on every push_data).
3. Publish. The event name in src/main.py (DATASET_ITEM_EVENT) must match the Console exactly.

Do not set prices in actor.json — Apify reads pay-per-event prices from the Console only.