Image Generation
Generate images with the images_generate tool. It handles text-to-image,
image-to-image (pass reference_images), and multi-image composition. By default
(model="auto") it picks the best model for the request; set model to choose one.
Requirements
This skill assumes the Hyper MCP is connected to your
agent so the images_generate tool is available. For brand-consistent ad creative
work, Firecrawl must also be configured under your Hyper integrations.
Call shape
images_generate(
requests=[{"id": "ad1", "prompt": "A polished SaaS ad, clean composition"}],
aspect_ratio="16:9", # "1:1" (default), "9:16", "16:9", "4:5", "2:3", "3:2", "3:4", "4:3", "21:9", ...
quality="standard", # "draft" | "standard" | "high"
n=1, # 1-4 images per request
model="auto", # see "Choosing a model" below
)
- Image-to-image / brand references: put files in the request:
requests=[{"prompt": "Compose into a gift basket", "reference_images": ["file1", "file2"]}].
- Reproducible output: pass
seed=....
- Ground in real-world search: pass
use_search=True.
- Do not display image URLs — they render automatically in chat.
Choosing a model
model="auto" is the right default. Override only when the task clearly calls for a
specific model:
| Task |
model |
| First-pass concepts / quick ad ideation |
gpt-image-2 |
| Image-to-image with references, high-resolution refinement, broad aspect ratios |
nano-banana |
| Readable text inside the image (posters, labels, infographics) or search-grounded scenes |
nano-banana-pro |
| Product photography, material/fabric fidelity, accurate spatial depth |
seedream-4.5 |
See references/image-prompting.md for per-model
prompt-writing tips.
Branded / website ad creatives — extract branding first
If the user gives a website URL and wants on-brand creatives:
- Call
firecrawl_branding_extract with the URL → returns brand colors, fonts,
personality/tone, and saved image files (logo, favicon, og_image).
- Optionally
firecrawl_urls_scrape with formats=["screenshot"] for visual context.
- Write the prompt using the actual hex colors, font names, and tone, and pass the
logo
file_id in reference_images.
The branding result's file field is a JSON data file, NOT an image — never pass it
as a reference. Only logo.file_id and images.*.file_id are usable images.
Higher-level workflows
For multi-shot product or marketplace work, prefer the workflow tools — they preserve
product identity and return structured results:
images_product_photoshoots_create — multi-shot product photography (studio,
lifestyle, hero, carousel, ad pack). See
references/product-photoshoot.md.
images_marketplace_cards_create — marketplace listing image sets (Amazon main +
secondary, A+ modules, Shopify). See
references/marketplace-cards.md.
Reminders
- Do NOT display image URLs to the user — they show automatically in chat.
- Refine vague prompts unless the user wants verbatim generation.
- Match
aspect_ratio to intent (social, print, web).
- Use
quality="high" for production, "draft"/"standard" while iterating.
- Generated
file_ids can be reused as reference_images in later calls.
- For website brand work, call
firecrawl_branding_extract before generating.
1---2name: image-generation3description: Generate images through the Hyper MCP with the unified `images_generate` tool — text-to-image, image-to-image, and branded ad creatives — choosing the model (gpt-image-2, nano-banana, nano-banana-pro, seedream-4.5) per task. Use when the user asks to generate an image, create an ad creative, do an image-to-image edit, render text inside an image, or produce a print-quality poster.4---56# Image Generation78Generate images with the `images_generate` tool. It handles text-to-image,9image-to-image (pass `reference_images`), and multi-image composition. By default10(`model="auto"`) it picks the best model for the request; set `model` to choose one.1112## Requirements1314This skill assumes the [Hyper MCP](https://app.hyperfx.ai/mcp) is connected to your15agent so the `images_generate` tool is available. For brand-consistent ad creative16work, Firecrawl must also be configured under your Hyper integrations.1718## Call shape1920```python21images_generate(22 requests=[{"id": "ad1", "prompt": "A polished SaaS ad, clean composition"}],23 aspect_ratio="16:9", # "1:1" (default), "9:16", "16:9", "4:5", "2:3", "3:2", "3:4", "4:3", "21:9", ...24 quality="standard", # "draft" | "standard" | "high"25 n=1, # 1-4 images per request26 model="auto", # see "Choosing a model" below27)28```2930- **Image-to-image / brand references:** put files in the request:31 `requests=[{"prompt": "Compose into a gift basket", "reference_images": ["file1", "file2"]}]`.32- **Reproducible output:** pass `seed=...`.33- **Ground in real-world search:** pass `use_search=True`.34- Do not display image URLs — they render automatically in chat.3536## Choosing a model3738`model="auto"` is the right default. Override only when the task clearly calls for a39specific model:4041| Task | `model` |42|------|---------|43| First-pass concepts / quick ad ideation | `gpt-image-2` |44| Image-to-image with references, high-resolution refinement, broad aspect ratios | `nano-banana` |45| Readable text inside the image (posters, labels, infographics) or search-grounded scenes | `nano-banana-pro` |46| Product photography, material/fabric fidelity, accurate spatial depth | `seedream-4.5` |4748See [references/image-prompting.md](references/image-prompting.md) for per-model49prompt-writing tips.5051## Branded / website ad creatives — extract branding first5253If the user gives a website URL and wants on-brand creatives:54551. Call `firecrawl_branding_extract` with the URL → returns brand colors, fonts,56 personality/tone, and saved image files (logo, favicon, og_image).572. Optionally `firecrawl_urls_scrape` with `formats=["screenshot"]` for visual context.583. Write the prompt using the actual hex colors, font names, and tone, and pass the59 logo `file_id` in `reference_images`.6061The branding result's `file` field is a JSON data file, NOT an image — never pass it62as a reference. Only `logo.file_id` and `images.*.file_id` are usable images.6364## Higher-level workflows6566For multi-shot product or marketplace work, prefer the workflow tools — they preserve67product identity and return structured results:6869- `images_product_photoshoots_create` — multi-shot product photography (studio,70 lifestyle, hero, carousel, ad pack). See71 [references/product-photoshoot.md](references/product-photoshoot.md).72- `images_marketplace_cards_create` — marketplace listing image sets (Amazon main +73 secondary, A+ modules, Shopify). See74 [references/marketplace-cards.md](references/marketplace-cards.md).7576## Reminders7778- Do NOT display image URLs to the user — they show automatically in chat.79- Refine vague prompts unless the user wants verbatim generation.80- Match `aspect_ratio` to intent (social, print, web).81- Use `quality="high"` for production, `"draft"`/`"standard"` while iterating.82- Generated `file_id`s can be reused as `reference_images` in later calls.83- For website brand work, call `firecrawl_branding_extract` before generating.