Skip to content

Choose a different AI model for every piece

No single image model is best at everything. artCart.AI connects to the major generators so you can match the model to the artwork instead of fighting the one you happen to be locked into.

You do not need an account with any of these providers, and there is no API key to configure. Pick a model at generation time and pay in credits.

Which model should I use?

Painterly, atmospheric work

Midjourney has a distinctive aesthetic that flatters landscapes, portraits and anything romantic. It returns four variations per run, which is useful when you are exploring rather than executing.

Artwork containing words

Ideogram and GPT Image render legible typography. If the piece is a poster, a quote print or anything with a sign in it, start here.

Long, specific prompts

GPT Image follows detailed instructions closely. Use it when you know exactly what you want and need it honoured.

Photorealism

Gemini 3.1 Flash Image handles natural light, skin and landscape convincingly. Gemini 3 Pro Image pushes further when the result has to survive close inspection.

Large prints

FLUX 1.1 Pro holds fine detail better than most, which matters once you are printing at 24×36 rather than looking at a thumbnail.

Volume and experimentation

Stable Diffusion and FLUX Schnell cost the least per image, so they are the sensible choice when you are generating dozens of options.

Every provider we support

Availability is set by your platform administrator. Models marked Available now can be selected today.

OpenAI

Available now

The GPT Image family. Excellent prompt adherence and the most legible text in images.

ModelCreditsShapes
GPT Image 2.5 Flare 10 3 ratios
GPT Image 2.5 Sunburst 10 3 ratios
GPT Image 2 5 3 ratios
GPT Image 1.5 4 3 ratios
GPT Image 1 4 3 ratios
GPT Image 1 mini 2 3 ratios

Direct: results usually arrive in well under a minute.

Google AI (Gemini)

Available now

Gemini image models. Strong photorealism, wide aspect-ratio support and fast turnaround.

ModelCreditsShapes
Gemini 3 Pro Image 5 7 ratios
Gemini 3.1 Flash Image 3 7 ratios
Gemini 3.1 Flash Lite Image 2 7 ratios
Gemini 2.5 Flash Image 2 7 ratios

Direct: results usually arrive in well under a minute.

Stability AI

Available now

Stable Image Ultra, Core and SD 3.5. Broad artistic range at a low per-image cost.

ModelCreditsShapes
Stable Image Ultra 4 7 ratios
Stable Image Core 2 7 ratios
Stable Diffusion 3.5 3 7 ratios

Direct: results usually arrive in well under a minute.

Black Forest Labs (FLUX)

Available now

FLUX 1.1 Pro and FLUX Dev. Very high detail; generation is queued and polled.

ModelCreditsShapes
FLUX 1.1 Pro 4 7 ratios
FLUX Pro 3 7 ratios
FLUX Dev 2 7 ratios

Queued: results arrive when the model finishes, typically under a few minutes.

PiAPI (Midjourney)

Midjourney through PiAPI. Returns a grid of four images per run.

ModelCreditsShapes
Midjourney Turbo 3 ×4 7 ratios
Midjourney Fast 2 ×4 7 ratios
Midjourney Relax 1 ×4 7 ratios

Queued: results arrive when the model finishes, typically under a few minutes.

Replicate

One key, hundreds of open models. Handy for SDXL, FLUX Schnell and Recraft.

ModelCreditsShapes
FLUX Schnell 1 7 ratios
FLUX Dev 2 7 ratios
Recraft V3 3 5 ratios
SD 3.5 Large 3 5 ratios
Upscale 2x 1 1 ratios
Upscale 4x 2 1 ratios

Queued: results arrive when the model finishes, typically under a few minutes.

xAI (Grok)

Grok image generation. OpenAI-compatible endpoint, up to ten images per run.

ModelCreditsShapes
Grok 2 Image 2 1 ratios

Direct: results usually arrive in well under a minute.

Ideogram

Best-in-class typography. Ideal for posters and any art with words in it.

ModelCreditsShapes
Ideogram V3 Quality 4 7 ratios
Ideogram V3 Balanced 2 7 ratios
Ideogram V3 Turbo 1 7 ratios

Direct: results usually arrive in well under a minute.

fal.ai

Fast hosted runtime for FLUX and friends. Queued API with polling.

ModelCreditsShapes
FLUX Schnell 1 5 ratios
FLUX Dev 2 5 ratios
FLUX 1.1 Pro 4 5 ratios

Queued: results arrive when the model finishes, typically under a few minutes.

Credit costs are the platform defaults and may be adjusted by your administrator. Where a model shows a multiplier, it always returns that many images per run.

Questions about models

Do I need my own API key for these?

No. artCart.AI holds the provider connections. You choose a model and spend credits — there is no account to open with OpenAI, Google or anyone else.

Can I switch models halfway through a project?

Yes, every run is independent. Many artists explore with a cheap model and then re-run their best prompt on a more expensive one for the final piece.

Why do some models cost more credits?

The credit price reflects what the provider charges us. Faster queues and higher-quality models cost more per image at the source.

Can I sell what these models produce?

Generally yes, but each provider sets its own commercial-use terms and they do change. Check the terms for the specific model before building a product line on it.

Which model is best for printing large?

FLUX 1.1 Pro and Gemini 3 Pro Image hold the most fine detail. See the model selection guide for a fuller comparison.

Try several models with your free credits