How to Use ImagineArt 1.5: 2K Photoreal Images at 6 Credits Each

ImagineArt 1.5 is VdoBloom’s 2K hyper-realistic text-to-image model, costing a flat 6 credits ($0.30) per image regardless of aspect ratio, with nine ratios available including the 3:1 and 1:3 panoramas most generators do not offer. It runs through Runware, it is tagged as a flexible-filter model rather than a strict one, and its stated specialities are photoreal output and accurate typography inside the image. This guide covers what it actually earns its keep on, the exact cost of every route through it, how to prompt it, and the two sibling models that beat it in specific situations.

What ImagineArt 1.5 is best at

Two things, and it is worth being precise about them because they determine whether this model belongs in your workflow at all.

The first is photorealism at 2K. The model is tuned for hyper-realistic output rather than illustration or stylised art. Product shots on a surface, portraits, interiors, food, architectural exteriors — the categories where “this looks like a photograph” is the pass condition. It is not the model you reach for when you want a painterly or graphic-novel look.

The second is text inside the image. Precise typography is part of its stated design, which matters more than it sounds. Most image models still mangle a sign, a label, a book spine or a packaging mock-up into approximate letterforms. If your brief includes a product label that has to read correctly, or a poster mock-up with a legible headline, that requirement alone narrows the catalog to a handful of models, and this is one of the cheaper entries in that group.

The third thing worth knowing is a format quirk: the ratio list includes 3:1 and 1:3. Those are extreme panoramas — website hero banners, email headers, tall sidebar rails — and generating them natively beats generating 16:9 and cropping, because a crop throws away the composition the model actually planned.

The real specs

• Cost: 6 credits per image, flat
• Resolution: up to 2K
• Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 3:1, 1:3
• Inputs: text prompt only — there is no image input slot on this variant
• Content tier: flexible, the more permissive of the image tiers
• Catalog rating: 3.7

That 3.7 is honest rather than flattering, and it is the number to hold in mind against the sections below. ImagineArt 1.5 is a capable, specific, well-priced tool, not the highest-rated image model on the platform.

Exact credit costs, and how the family compares

A VdoBloom credit is five cents at the Lite tier, where $15 buys 300 credits, so the conversion below is direct. Credits bought as one-time packs start at $2.49 for 75 and never expire; the full table lives on the pricing page.

Model	Credits per image	Cost per image	Max resolution	Best for
ImagineArt 1.5	6	$0.30	2K	Photoreal work with legible text, panoramic ratios
ImagineArt 2.0	10	$0.50	2K (1K selectable)	Complex multi-part prompts, vivid colour
ImagineArt 1.5 Pro	12	$0.60	4K	Print and large-format output
FLUX.2 [dev]	4	$0.20	—	Cheap high-volume exploration
Riverflow 2.0 Fast	10	$0.50	2K	Fast turnaround on general work

Practical reading of that table: 50 images from ImagineArt 1.5 is 300 credits, exactly one month of the Lite plan. The same 50 images from ImagineArt 1.5 Pro is 600 credits. Which means the standard model buys you twice the number of attempts for the same money, and attempts are what image generation actually consumes.

When to pick a sibling instead

Three clean decision rules:

1. Going to print, or to a wall? Take 1.5 Pro. It renders natively at 4K rather than 2K and adds the 21:9 ultrawide ratio. Upscaling a 2K generation is not the same thing as generating at 4K, and at 12 credits the premium is six cents per image over standard.
2. Prompt with several interacting parts? Take ImagineArt 2.0. It is the reasoning-based model in the line, which is the difference that shows up when a prompt says three specific things must be true at once — a particular subject, in a particular position, holding a particular object. It also adds 4:5, 5:4 and 21:9, and lets you drop to 1K deliberately when you want speed.
3. Just exploring? Take something cheaper. FLUX.2 dev at 4 credits is a third less per image, and for the first twenty throwaway compositions that difference is the whole point. Move to ImagineArt 1.5 once you know what you are making.

How to prompt it well

ImagineArt 1.5 is text-only, so the prompt carries everything. Four things that reliably help on this model specifically:

Write the text you want in quotes. If typography is why you chose this model, spell out the exact string — a sign reading “Open Daily” — rather than describing it as a sign with opening hours. Short strings survive far better than long ones; aim for three or four words on a label, not a paragraph.

Use photographic vocabulary, not art vocabulary. The model is tuned for realism, so it responds to lens, distance and lighting terms. Describe the shot the way a photographer would brief it — 50mm, shallow depth of field, soft window light from the left — instead of asking for a beautiful or stunning image.

Pick the ratio before the prompt, not after. A composition written for 1:1 does not survive being regenerated at 3:1. If the deliverable is a banner, prompt for a banner: state where the subject sits in the frame and what fills the long empty run beside it.

Change one variable per attempt. At 6 credits an image, a disciplined run of eight single-variable attempts costs 48 credits and teaches you something. Eight scattered rewrites cost the same and teach you nothing.

A worked example

Brief: a website hero banner for a small-batch coffee roaster, panoramic, with the product name legible on the bag.

Setup: ratio 3:1, model ImagineArt 1.5. Prompt: A kraft-paper coffee bag standing on a pale concrete surface, label reading “Slow Roast” in bold sans-serif, shot on 50mm at f/2, soft diffuse daylight from the right, subject positioned left of centre, long empty concrete run extending to the right of frame.

Cost: 6 credits ($0.30) per attempt. Budget six attempts to lock the label and the composition — 36 credits, $1.80 all in. That empty right-hand run is deliberate: it is where the headline and button go once the banner reaches the page, and asking for it in the prompt is what stops the model from centring the bag and leaving you nowhere to put type.

If it needs to go to print: re-run the winning prompt once on 1.5 Pro at 12 credits rather than upscaling the 2K file. One extra generation, six extra cents, native 4K.

To run any of this, open the image generation workspace and pick ImagineArt 1.5 from the model list, or read the catalog entry on the model page first. New accounts get 10 free credits with no card, which is one ImagineArt 1.5 image with change left over — enough to see the output quality before deciding anything. Paid plans download without a watermark and with commercial rights.

Frequently asked questions

How much does one ImagineArt 1.5 image cost?

6 credits, which is $0.30 at the Lite rate of $0.05 per credit. The price is flat — a 3:1 panorama costs the same as a 1:1 square.

Can ImagineArt 1.5 edit an existing image?

No. This variant takes a text prompt only and has no image input. For edits, use one of the dedicated image-editing models in the VdoBloom image workspace.

Does it really render readable text?

Accurate typography is part of what the model is built for, and it handles short strings noticeably better than general-purpose generators. It is not infallible — put the exact words in quotes, keep them short, and expect to run a few attempts on a label that must be perfect.

What is the difference between ImagineArt 1.5 and 2.0?

2.0 is the reasoning-based model at 10 credits, better at prompts with several conditions that all have to hold at once, and it adds the 4:5, 5:4 and 21:9 ratios plus a selectable 1K mode. 1.5 at 6 credits is the better value for straightforward photoreal shots.

What does the flexible content tier mean here?

It is the more permissive of the VdoBloom image tiers, so swimwear, dance, fitness and fashion subjects that stricter mainstream generators refuse will generate. Explicit and illegal content stays blocked on every model in the catalog.