Test log
ImagineArt test data
Raw results for the ImagineArt review. The verdict is in the review.
- Tested
- 6 November 2026
- Tester
- Miles Foster
- Account
- Free demo account, signed up by RankDude, nothing paid to the brand
- Tests
- 10
- Screenshots
- 5
- Method
- Published methodology
- Environment
- Google Chrome on desktop
Results
| ID | Test | Key measurement | Score |
|---|---|---|---|
| T01 | Text in Images | Run: 0 of 6 | n/a |
| T02 | Style Range | Run: 0 of 10 | n/a |
| T03 | Character & Subject Consistency | Run: 0 of 12 | n/a |
| T04 | Editing, Inpainting & Image-to-Image | Run: 0 of 12 | n/a |
| T05 | Prompt Adherence | Elements correct: 12 of 12 | 5/5 |
| T06 | Photorealism & Anatomy | Images that pass: 1 of 2 | 2/5 |
| T07 | Resolution & Upscaling | Native size: 2,048 px (4.2 MP) | 3/5 |
| T08 | Speed per Image | Median per image: 36.2 s | 3/5 |
| T09 | Usage Rights, Watermarks & Privacy | Points met: 4 of 8 checked | 2/5 |
| T10 | Pricing & Value | Free images: 3 in all | 2/5 |
| Overall (in the review) | 3/5 | ||
Text in Images
n/aNot Tested- Method
- Six prompts with set wording.
- Run
- 0 of 6
- Remark
- Not graded: nothing was run.
Raw notes
Six prompts with set wording.
Not run: each image cost about 26 credits and the balance paid for three.
- Run on a plan with enough credits.
Style Range
n/aNot Tested- Method
- Ten style prompts for one lighthouse.
- Run
- 0 of 10
- Remark
- Not graded: nothing was run.
Raw notes
Ten style prompts for one lighthouse.
Not run: ten images would cost about 260 credits.
- Run on a plan with enough credits.
Character & Subject Consistency
n/aNot Tested- Method
- A described woman and a described thermos in six scenes each.
- Run
- 0 of 12
- Remark
- Not graded: nothing was run.
Raw notes
A described woman and a described thermos in six scenes each.
Not run: twelve images would cost about 310 credits.
- Run on a plan with enough credits.
Editing, Inpainting & Image-to-Image
n/aNot Tested- Method
- Twelve fixed edits.
- Run
- 0 of 12
- Remark
- Not graded: nothing was run.
Raw notes
Twelve fixed edits.
Not run: no credits were left for edits.
- Run on a plan with enough credits.
Prompt Adherence
5/5Tested- Method
- Two of the 40 fixed prompts, the portraits P1 and P2, typed in by script.
- Elements correct
- 12 of 12
- Prompts run
- 2 of 40
- Remark
- Scale 5: 90 percent or more of elements correct. The sample is 2 prompts and 12 elements, not 40 prompts and about 240.
Raw notes
Two of the 40 fixed prompts, the portraits P1 and P2, typed in by script. The first image of each was kept and every written element was counted at full size.
12 of 12 elements were right (100 percent). The elderly man had a white beard, round glasses, a blue wool cardigan, sat on a park bench and stood in soft morning light. The young woman laughed, with short black hair, freckles, a yellow raincoat and a blurred city street.
Both prompts were portraits. The busy scenes, where other generators miscount objects, were not run.
- 2 of 40 prompts, both portraits. The score would likely fall on busy scenes.
Photorealism & Anatomy
2/5Tested- Method
- The two portrait images, one run each, inspected at 100 percent for faults that give a generated image away.
- Images that pass
- 1 of 2
- Hand and face faults
- 0
- Remark
- Scale 2: 7 to 12 of 24 pass. One of two (50 percent) stands for 12 of 24.
Raw notes
The two portrait images, one run each, inspected at 100 percent for faults that give a generated image away.
Passed (1): the man on the bench, with natural skin, eyes, teeth and two correct hands.
Failed (1): the woman's face, hair and raincoat are convincing, but a translucent X shape runs through the blurred street behind her on the left, with a second faint diagonal band, and it does not belong to any object in the scene.
Hand faults 0. Face faults 0. Background faults 1.
- Two images in one run, not 24 images in two runs. One flaw in a blurred background decides the grade.
Resolution & Upscaling
3/5Tested- Method
- The saved images were measured, the app's resolution setting was read, and the 2x Imagine Upscalar was run on the woman's image after its price dialog was read.
- Native size
- 2,048 px (4.2 MP)
- Upscale result
- None in 14 min
- Remark
- Scale 3: 1 megapixel or more native with an upscaler that did not deliver. The native size would meet scale 5, but its clean upscale to 4,096 pixels could not be shown.
Raw notes
The saved images were measured, the app's resolution setting was read, and the 2x Imagine Upscalar was run on the woman's image after its price dialog was read.
Both saved images are 2,048 by 2,048 pixels (4.2 megapixels) as JPEG: 2,190,265 and 1,811,101 bytes. The settings dialog lists Auto, 1K, 2K and 4K.
The upscale dialog offers 2x, 4x, 8x and 16x and shows "Generation will cost 8 credits". After I pressed Generate at 2x the page showed a spinner for 14 minutes and never returned an image. The balance stayed at 23.
With no upscaled file there was nothing to compare for invented texture.
- One upscale attempt, and it did not finish. A retry might behave differently.
Speed per Image
3/5Tested- Method
- Each generation was timed from pressing Generate to the full-size image on screen.
- Median per image
- 36.2 s
- Images timed
- 2
- Remark
- Scale 3: a median of 20 to 39.9 seconds.
Raw notes
Each generation was timed from pressing Generate to the full-size image on screen.
The two timed images took 36.3 and 36.0 seconds, a median of 36.2 seconds.
Auto picked ImagineArt 2.0 for both. Nothing failed, and a sale pop-up opened over the page after each image.
- Two timings, not 40, on one day.
Usage Rights, Watermarks & Privacy
2/5Tested- Method
- The terms of service (read 6 November 2026) and the saved files were checked against the ten-point list.
- Points met
- 4 of 8 checked
- Remark
- Scale 2: 4 or 5 of 10. 4 of 8 checked is 50 percent, so about 5 of 10.
Raw notes
The terms of service (read 6 November 2026) and the saved files were checked against the ten-point list. Eight points could be checked on the free plan.
Met (4): paid subscribers may use outputs commercially; the terms say who may use output (ImagineArt owns the service and content except the outputs it licenses to you); the two saved 2,048 px files show no visible watermark; the terms say ImagineArt "will not use your personal content for AI model training without your consent".
Not met (4): free plan use "may be limited to non-commercial, personal purposes"; free generations are public, since Private Generations is crossed out on the cheapest paid plan; no stated source for the training data; no content credentials marker (c2pa) found in either file.
Not tested (2): a watermark on paid downloads, and whether real people and brands are refused when asked for.
- Terms read by me, not a lawyer. The free download button was not compared; files were saved from the viewer.
Pricing & Value
2/5Tested- Method
- The free plan was used until three images were made, the credit balance was read before and after each run, and the plans page was read in monthly and yearly billing.
- Free images
- 3 in all
- Cost per image on Basic
- about $0.11
- Remark
- Scale 2: $0.10 or more an image, and the credit count on the page does not match the real cost.
Raw notes
The free plan was used until three images were made, the credit balance was read before and after each run, and the plans page was read in monthly and yearly billing.
Free: 100 credits, enough for three images at about 26 credits each (51 for two images, 26 for one).
The cheapest paid plan, Basic, is $13 a month ($13 a month paid yearly) for 3,000 credits. At 26 credits an image that is about $0.11 an image billed monthly and $0.08 billed yearly.
The plans page says Basic is "≈ 600 images or 96 videos", which is 5 credits an image. The images I made cost five times that, so the page's count did not match what I paid.
- Cost per image is for the Auto model on a portrait prompt; cheaper models cost fewer credits. No plan was bought.
Coverage
Tested
- 2 of the 40 fixed prompts (an elderly man on a bench and a laughing young woman), counted at full size
- Time and credit cost of each generation
- The 2x Imagine Upscalar on one image, with its price dialog
- The terms of service read in full, and the plans page read in monthly and yearly billing
- The size and file of each saved image
Not tested or limited
- The free plan's 100 credits paid for three images: 2 of the 40 prompts were run, so every grade rests on a sample of two
- The 2x upscale ran for 14 minutes without returning an image, so no upscaled file was compared
- Text in images, style range, consistency and editing were not run: each needs 6 to 12 more images
- No plan was bought: paid downloads, private generations and cancellation were not tested
- Prompts naming real people or brands were not run, so that point of the rights checklist was not tested
Still to test
- Text, style range, consistency and editing on a paid plan
- Busy scenes and the other 38 prompts
- The 2x upscale on a second try
- Paid downloads, private generations and cancellation after buying Basic