Cycling through all six. Tap any point to stop.

Measured on six things

AI image and video generator with all the best models in ...

Accessibility

Lighthouse a11y 74/100 with real failures: unlabeled buttons and form elements, low-contrast text, links without discernible names, and a viewport that blocks pinch-zoom on mobile. No accessibility statement found.

Screenshots

1 / 1

ImageGPT

About this app

ImageGPT is a web-based AI image and video generator that lets you create and edit visuals using a range of leading AI models from a single interface. You describe what you want, and it handles logos, headshots, product photos, thumbnails, background removal, and video editing. It stands out for multi-model access, a no-subscription credit model, and a template-driven workflow, making it a flexible pick for creators, marketers, and small businesses who want pro-level results without juggling several tools.

Key features

1Multi-model access: text-to-image, image-to-video, background removal, upscaling, vectorizing, and video download in one chat interface.
2Pay-as-you-go credit packs from $5.99 (25 credits) to $149.99 (2,000 credits) with no subscription and credits that never expire.
3Auto model routing plus manual override, with parallel task execution and multi-step workflow support documented in the docs.

Target customer profile

Drawn from the product itself, not from a survey.

Demographic

Social media managers and content creators

Pain points

Tight turnaround times, inconsistent visual quality across platforms

Primary needs

Fast thumbnails, background removal, and model variety in one tool

Demographic

Small business owners and e-commerce sellers

Pain points

No budget for design agencies, poor product photo quality

Primary needs

Affordable product photos, logos, and pay-as-you-go pricing

Demographic

Freelance designers and marketers

Pain points

Paying for multiple AI tools and subscriptions

Primary needs

Access to several top models and image/video editing in one place

Scout Review

Written by AI from measured evidence, scored out of 100.

ImageGPT is a competent multi-model AI image and video generator wrapped in a chat interface that anyone can use on day one. The docs are thorough, desktop performance is strong, and the pay-as-you-go credit model with non-expiring credits is a genuine differentiator against subscription-heavy competitors. The weaknesses are concrete: mobile LCP at 4.6s, a security header checklist at 5/13, and an accessibility score of 74 with unlabeled buttons and a zoom-blocking viewport. The homepage also fails to show a single example output — a strange omission for a visual product. It's a solid tool for creators who already know what they want; it's a harder sell for anyone who needs proof before paying.

Measure by measure

Usability

Chat interface plus auto model routing keeps onboarding near-zero for anyone who's used ChatGPT. Docs are genuinely thorough — 12 model categories, tools, and a Discord support path. No free trial, only 'earn free credits'.

Design

A deliberately bare chat-first UI that borrows ChatGPT's layout wholesale. Clean and functional, but the homepage shows no product screenshots, no demo, and no visual proof of output quality — a miss for an image generator.

Speed

Desktop Lighthouse 94/100 is genuinely snappy. Mobile drops to 75 with a 4.6s LCP — the largest contentful paint is slow enough to feel on a phone. CLS 0 and TBT 280ms are clean.

Security

HTTPS enforced, secure cookies, SPF and DMARC in place, Stripe handles payments. But no HSTS, no CSP, no frame protection, no Referrer-Policy — 5/13 header checklist points. No SOC 2, ISO, or security page found.

Accessibility

Growth

Multi-model aggregator with pay-as-you-go credits is a real wedge against Midjourney's Discord wall and Canva's subscription. But 'we route to the best model' is copyable, and the category is crowded.

Improvement areas

Homepage shows no example outputs or demo — critical trust gap for a visual product.
Mobile LCP at 4.6s and security headers at 5/13 (no HSTS, CSP, or frame protection).
Accessibility failures: unlabeled buttons/forms, low contrast, and a viewport that blocks pinch-zoom.

Design & Usability

ImageGPT's design is a conscious bet: strip everything down to a chat box and let the models do the talking. It works — anyone who has typed into ChatGPT knows exactly what to do — but for an image generator, showing zero example outputs on the homepage is a strange choice. The docs are the real design win: a clean sidebar covering text-to-image, image-to-video, background removal, upscaling, vectorizing, and video download, all documented. Usability is strong for the target user (creators who want speed), weaker for anyone who needs to see proof before paying. The absence of a free trial, replaced by 'earn free credits,' adds friction at exactly the moment trust matters most.

Speed & Security

Speed is a tale of two devices. Desktop Lighthouse hits 94/100 — genuinely fast. Mobile lands at 75 with a 4.6s LCP, which on a phone means the chat box takes nearly five seconds to become useful. CLS 0 and TBT 280ms show the layout is stable and the main thread isn't choking, so the mobile LCP is likely a hero asset or font issue rather than a structural problem. Security is thinner: HTTPS and secure cookies are there, Stripe handles payments, SPF and DMARC are configured — but the app ships without HSTS, CSP, frame protection, Referrer-Policy, or Permissions-Policy, landing at 5/13 on the header checklist. No SOC 2, ISO, or security page was found. For a tool handling user uploads and payment flows, that's a gap worth closing.

Accessibility & Growth

Accessibility is the weakest measured dimension: Lighthouse 74/100 with concrete failures — buttons without accessible names, form inputs without labels, low-contrast text, links without discernible names, and a viewport that disables pinch-zoom on mobile. For a chat-based product where the entire interaction is typing into a form, unlabeled form elements are a direct hit on the core flow. No accessibility statement was found. Growth is a more interesting question: the multi-model aggregator pitch is real — Midjourney lives in Discord, Firefly is locked to Adobe, Canva bundles AI into a subscription — and 'no subscription, credits never expire, all the top models in one chat' is a coherent wedge for SMBs and creators who don't want three subscriptions. But the wedge is copyable. The moment a competitor adds a second model or a web UI, 'we aggregate' stops being a reason to switch. The domain is 6.3 years old, suggesting a long-running operation, but the current positioning reads as a recent pivot. The idea can grow; the moat needs work.

Conclusion

If you're a social media manager or small business owner who already knows the prompt you want to run and just needs a fast, no-subscription way to run it across multiple models, ImageGPT is a reasonable pick — the docs are real, the credits don't expire, and desktop performance is snappy. If you're comparing against Midjourney's output quality or Canva's all-in-one workflow, the lack of visible examples on the homepage makes that comparison harder than it should be. The product is functional and the pricing model is honest; the polish and trust signals are where the work remains.

Suggestions for improvement

Add a gallery of real generated outputs to the homepage — an image generator with no images on its landing page is leaving trust on the table.
Fix the mobile LCP (4.6s): preload the hero font/asset and audit what's blocking first paint on phones.
Ship the missing security headers — HSTS, CSP, X-Content-Type-Options, Referrer-Policy, Permissions-Policy — and publish a short security page.
Fix the accessibility failures: label every button and form input, raise text contrast, and remove user-scalable=no from the viewport.

How we score

Alternatives comparison

Named competitors, point by point. Nobody paid to appear here or to be left out.

	ImageGPT	Midjourney	Canva	Adobe Firefly
Pricing model	Pay-as-you-go credits, $5.99–$149.99, never expire, no subscription	Monthly subscription tiers (Basic/Standard/Pro)	Freemium with Canva Pro subscription	Credit-based within Adobe Creative Cloud subscription
Interface	Web-based chat interface with auto model routing	Primarily Discord bot, with a newer web interface	Full drag-and-drop design canvas with AI tools	Web app plus integration inside Photoshop/Illustrator
Model variety	Multiple third-party models for image, video, audio, and editing	Proprietary Midjourney models only	Canva's own AI plus some partner models	Adobe Firefly models only
Video generation	Text-to-video, image-to-video, video-to-video, clip combining	Limited video features (recently added)	Video editing with AI, limited generative video	Firefly video generation in beta
Background removal / editing	Background removal, upscaling, vectorizing, format conversion	Not a core feature	Built-in background remover and full editor	Generative fill and removal in Photoshop