AI image generation

Overview

Your app can generate images using state-of-the-art models for product mockups, blog cover art, social media graphics, profile avatars, and more. Images are created in the background — you ask for one and it appears a few seconds later, so the screen never freezes while it's working.

What it supports

• Prompt-to-image generation — describe an image in plain text and get back a generated image stored in your app
• Choice of models — use a curated set of leading models (including OpenAI GPT Image 2 and the FLUX family), and decide which ones your app offers
• Full control of the look — shape, level of detail, quality, and how many images to create
• Edit existing images — refine or create variations from an uploaded image, not just from text
• Consistent styles — set a house style (e.g., photorealistic, illustrated, flat icon) so every generated image looks cohesive
• Asynchronous jobs — generation runs in the background; your app polls or subscribes for completion so the UI stays responsive
• Activity dashboard — see recent jobs, their status, and cost-per-image inside the app's Settings tab
• Auto-storage — finished images land in a dedicated generated-images storage category and can be referenced anywhere in your app

Available models

Pentoggle gives your app a curated set of leading image models. You choose which ones your app uses — and can offer more than one so the right model fits each job.

• OpenAI GPT Image 2 — best for text-in-image, infographics, and precise layouts
• FLUX.1 Schnell — fastest, great for drafts and quick iteration
• FLUX.1 Dev — balanced everyday quality
• FLUX.1 Pro v1.1 & Pro v1.1 Ultra — high-end, photoreal, marketing-grade art
• FLUX.1 Kontext Pro — image editing and variations (image-to-image)
• Nano Banana Pro — high quality, supports up to 4K
• Recraft V3 — clean brand/vector-style graphics
• Ideogram V2 — strong typography and text-heavy designs
• Stable Diffusion 3 Medium — general-purpose generation

Each model's cost per image is shown when you set it up, so you can balance quality against spend. Pentoggle adds new models over time.

What you can control

You don't set any of this up by hand — just describe what you want when you build the feature, and optionally let your own users choose. Every image can be tailored by:

• Shape — square, portrait, landscape, or ultrawide
• Detail — from standard up to high-resolution (4K) for print-quality output
• Quality — quick drafts while iterating, or top quality for finished art (higher quality and detail cost a little more)
• Number of images — create up to four at once so users can pick their favourite
• Editing an existing image — start from an uploaded picture to tweak it or make variations, instead of creating from scratch
• A consistent look — keep every image on-brand with a house style you set once
• A spending limit — cap how much a single image can cost, so usage never surprises you

If a chosen model can't do exactly what you asked (a shape or size it doesn't offer), Pentoggle automatically uses the closest option so the image still comes out.

Example prompts

"Add a 'Generate cover image' button on each blog post
that creates a hero image from the post title."

"Let users describe their dream product and generate three
photorealistic mockups they can choose from."

"Auto-generate a square OG image for every page on my
site using the page title and a brand-consistent style."

Setup (none required)

Image generation is fully managed by Pentoggle. There's no provider account to create and nothing technical to wire up — just ask the agent to add image generation, then choose which models to offer and the style you want.

Image Generation settings

When your app uses image generation, an Image Generation card appears in your app's Settings tab. From there you can:

• Choose which models your app uses — each shown with its cost per image
• Set a spending limit per image — so a single generation never costs more than you allow
• Set a consistent style — a house style applied to every image, with a live preview
• Watch the activity dashboard — recent images, their status, and credits used

How it works

1. Something in your app (a button click, a scheduled task, etc.) asks for an image from a description, along with any options you've enabled — like shape, quality, or an image to edit
2. Pentoggle applies your chosen style and sends the request to the best-matched model
3. Your app immediately shows a “generating…” state, so nothing blocks while it works
4. The image is created and saved into your app's file storage automatically
5. As soon as it's ready, your app displays it — no manual steps
6. The image, its prompt, and its cost are recorded in the activity dashboard for later review

Where the images live

Generated images are saved automatically into your app's file storage. You can show them anywhere in your app, browse past generations, and delete old ones — just like any other file.

Limitations

• Models come from a curated set — you choose which of Pentoggle's models your app uses, but you can't bring an outside model of your own
• Shapes are a fixed set — square, portrait, landscape, and ultrawide; a very specific ratio (like 4:5) may be adjusted to the closest available shape
• Not instant — images take a few seconds and appear in the background, so your app shows a brief “generating” state rather than freezing
• Generation cost is metered — image generation consumes your app's live capacity, so high-volume usage will register on the usage dashboard
• Content policy applies — requests that violate the model's content policy will fail with a clear message; the agent can guide you on safer phrasing

Troubleshooting

• An image is taking too long — check the activity dashboard for an error; occasional hiccups are retried automatically
• Images don't match your style — adjust the style in Settings, then ask the agent to “refine the image style” from chat for fuller control
• An image hasn't appeared yet — it may still be finishing; give it a moment and it'll show up automatically
• Costs spiking unexpectedly — check the activity dashboard for high-volume callers (often a misconfigured loop) and rate-limit the calling code path