Technical

Technical diagrams, blueprints, formulas

5 tasks · 11 models tested · 55 results

Musical score

image
google gemini-2.5-flash-image
4.9/10 6.5 s
google gemini-2.5-flash-image
Cost < 0.01 $
Resolution 1344 x 768
Time 6.5 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.88
Review
The image possesses a compelling visual aesthetic that mimics the look of an old engraving, but it fails fundamentally on a functional level. The musical symbols and notation are completely incoherent and illegible, which contradicts the requirement for 'faithful notation.' The model has prioritized the appearance of the score at the expense of its actual structure.
google gemini-3-pro-image-preview
4.9/10 51.9 s
google gemini-3-pro-image-preview
Cost < 0.01 $
Resolution 2752 x 1536
Time 51.9 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.88
Review
The image is aesthetically successful and adheres to the "professional engraving" style, but it fails completely on its primary function: musical notation. The symbols and notes are incoherent visual artifacts that do not constitute a usable score, which contradicts the request for "faithful notation."
google imagen-4.0-fast-generate-001
4.2 s
google imagen-4.0-fast-generate-001
Cost 0.02 $
Resolution 1408 x 768
Time 4.2 s
google imagen-4.0-generate-001
3.9/10 8.8 s
google imagen-4.0-generate-001
Cost 0.04 $
Resolution 1408 x 768
Time 8.8 s
Matania Judgment
Quality
6
Composition
7
Creativity
5
Text accuracy
1
Fidelity
3
Overall
3.88
Review
The image fundamentally fails in its primary mission: the musical notation is completely illegible and inconsistent, failing to form an actual score and instead presenting a visual imitation of abstract symbols. While the engraving's aesthetics are respected, the "fidelity" is very low because the model does not produce usable or even semi-realistic music, but merely artifacts resembling notes.
google imagen-4.0-ultra-generate-001
3.9/10 12.3 s
google imagen-4.0-ultra-generate-001
Cost 0.08 $
Resolution 1408 x 768
Time 12.3 s
Matania Judgment
Quality
6
Composition
7
Creativity
5
Text accuracy
1
Fidelity
3
Overall
3.88
Review
The image fundamentally fails in its primary task: although it aesthetically resembles an ancient score, the musical symbols are incoherent and illegible artifacts that do not constitute actual notation. Fidelity is very low because the model has produced a visual imitation devoid of musical substance rather than a "faithful notation."
openai chatgpt-image-latest
4.3/10 44.7 s
openai chatgpt-image-latest
Cost 0.22 $
Resolution 1536 x 1024
Time 44.7 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
1
Fidelity
3
Overall
4.25
Review
The image is aesthetically pleasing and resembles an ancient musical score, but it fails completely on a technical level. The notes and musical symbols are incoherent visual artifacts that follow no real notation rules, making the "notational fidelity" (which was explicitly requested) almost non-existent.
segmind ideogram-3
4.5/10 15.6 s
segmind ideogram-3
Cost 0.04 $
Resolution 1344 x 768
Time 15.6 s
Matania Judgment
Quality
6
Composition
7
Creativity
5
Text accuracy
2
Fidelity
4
Overall
4.5
Review
The image succeeds in capturing the aesthetic look of an old engraved score, but fails completely on its primary function: the musical notation. The symbols are incoherent visual artifacts that follow no rules of music theory, making the requested "faithful notation" impossible. Fidelity is low because the model prioritizes the object's appearance over the technical content of the score.
segmind seedream-4.5
3.1/10 16.5 s
segmind seedream-4.5
Cost 0.04 $
Resolution 2560 x 1472
Time 16.5 s
Matania Judgment
Quality
4
Composition
5
Creativity
3
Text accuracy
1
Fidelity
3
Overall
3.13
Review
Technically, the image fails to represent an actual musical score; the musical symbols are incoherent and illegible visual artifacts. While the aesthetic evokes an old engraving, prompt fidelity is very low because there is no "faithful" or usable musical notation, only an abstract visual imitation of music.
segmind seedream-v5-lite
3.6/10 31.6 s
segmind seedream-v5-lite
Cost 0.04 $
Resolution 2848 x 1600
Time 31.6 s
Matania Judgment
Quality
4
Composition
5
Creativity
3
Text accuracy
1
Fidelity
4
Overall
3.63
Review
The image fails on the functional aspect of the score: the musical symbols are illegible artifacts and do not constitute actual notation. While the aesthetic evokes old engraved paper, prompt fidelity is low because this is not "faithful notation," but rather visual chaos imitating the form of music.
xai grok-imagine-image
4.8/10 7.5 s
xai grok-imagine-image
Cost 0.02 $
Resolution 2752 x 1504
Time 7.5 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
1
Fidelity
4
Overall
4.75
Review
The image presents a convincing visual aesthetic of an ancient score, but fails completely on its primary function: musical notation. The symbols are abstract, illegible artifacts that do not adhere to the rules of music theory, making the requested "faithful notation" impossible. Fidelity is low because the model prioritizes aesthetic appearance over the technical accuracy of the requested object.
xai grok-imagine-image-pro
4.9/10 19.4 s
xai grok-imagine-image-pro
Cost 0.07 $
Resolution 2816 x 1536
Time 19.4 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.88
Review
The image is aesthetically successful and visually resembles an ancient musical score, but it fails completely on a functional level. The musical notation is composed of abstract and incoherent symbols that follow no real reading rules (very low text_accuracy). Although the 'engraved' style is maintained, the promise of 'faithful notation' is not met, which heavily impacts prompt adherence.

Calligraphy 5 scripts

image
google gemini-2.5-flash-image
6.5/10 7.6 s
google gemini-2.5-flash-image
Cost < 0.01 $
Resolution 1344 x 768
Time 7.6 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The image is aesthetically very successful, boasting excellent technical quality and a harmonious composition. However, textual fidelity and precision are problematic: although the stylistic intent is present, the calligraphic characters (notably Arabic and Chinese) do not necessarily form the word 'PEACE' in a legible or correct manner, and the execution of certain scripts appears to be a visual approximation rather than true, accurate scriptural calligraphy.
google gemini-3-pro-image-preview
5.0/10 47.1 s
google gemini-3-pro-image-preview
Cost < 0.01 $
Resolution 2752 x 1536
Time 47.1 s
Matania Judgment
Quality
8
Composition
7
Creativity
7
Text accuracy
2
Fidelity
4
Overall
5
Review
The model fails significantly on textual accuracy and prompt adherence: it is unable to correctly represent the five requested scripts and mixes styles incoherently. While the visual aesthetics are acceptable, the semantic content is incorrect, which is critical for a calligraphy task.
google imagen-4.0-fast-generate-001
3.3 s
google imagen-4.0-fast-generate-001
Cost 0.02 $
Resolution 1408 x 768
Time 3.3 s
google imagen-4.0-generate-001
5.3/10 8.8 s
google imagen-4.0-generate-001
Cost 0.04 $
Resolution 1408 x 768
Time 8.8 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
2
Fidelity
4
Overall
5.25
Review
The image is aesthetically very successful, boasting excellent technical quality and a balanced composition. However, the model fails heavily on textual accuracy: the characters shown do not match the requested scripts (the 'Chinese' is an abstract scribble, the 'Japanese' is incoherent) and the concept of translating the word 'PEACE' into these systems is poorly executed. Fidelity is low as the specific linguistic constraints are not respected.
google imagen-4.0-ultra-generate-001
5.4/10 15.9 s
google imagen-4.0-ultra-generate-001
Cost 0.08 $
Resolution 1408 x 768
Time 15.9 s
Matania Judgment
Quality
8
Composition
9
Creativity
8
Text accuracy
2
Fidelity
4
Overall
5.38
Review
The image is aesthetically very successful, boasting excellent visual quality and a harmonious composition. However, the model fails almost entirely on textual accuracy and fidelity: the scripts do not match the requested languages (Chinese is replaced by abstract or incorrect forms, and Japanese/Hindi are poorly rendered) and the text is illegible, such as the word 'PEACE'. The fidelity score is severely impacted by the inability to respect the specific constraint of the 5 precise scripts.
openai chatgpt-image-latest
5.3/10 37.3 s
openai chatgpt-image-latest
Cost 0.21 $
Resolution 1536 x 1024
Time 37.3 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
2
Fidelity
4
Overall
5.25
Review
The image is aesthetically very successful, featuring beautiful composition and high graphical quality. However, the model fails heavily on textual accuracy and fidelity: it fails to write the word 'PEACE' consistently across the five requested scripts, mixing glyphs that do not always match the specified languages or creating unreadable hybrid forms.
segmind ideogram-3
8.5/10 13.4 s
segmind ideogram-3
Cost 0.04 $
Resolution 1344 x 768
Time 13.4 s
Matania Judgment
Quality
8
Composition
9
Creativity
8
Text accuracy
7
Fidelity
9
Overall
8.5
Review
The image is aesthetically very successful, featuring a balanced composition and high rendering quality. The model adheres well to the instruction for five different scripts, although the calligraphic precision of certain characters (notably the Chinese and Arabic) might be subject to debate by a linguistic expert. Prompt fidelity is excellent.
segmind seedream-4.5
4.5/10 11.7 s
segmind seedream-4.5
Cost 0.04 $
Resolution 2560 x 1472
Time 11.7 s
Matania Judgment
Quality
6
Composition
5
Creativity
7
Text accuracy
2
Fidelity
4
Overall
4.5
Review
The model fails significantly on textual accuracy (text_accuracy), producing glyphs that either do not match the requested scripts or are unreadable. While the overall aesthetics are correct, fidelity is very low because the writing systems (notably Arabic and Hindi) are misinterpreted or nonexistent, failing to adhere to the instruction for 5 specific scripts.
segmind seedream-v5-lite
4.4/10 36.7 s
segmind seedream-v5-lite
Cost 0.04 $
Resolution 2848 x 1600
Time 36.7 s
Matania Judgment
Quality
6
Composition
5
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.38
Review
The image fails significantly in terms of textual accuracy and prompt adherence. While the calligraphic aesthetic is present, the characters do not match the requested scripts (a mix of abstract shapes and incoherent glyphs), and the word 'PEACE' is not correctly rendered in the specified languages.
xai grok-imagine-image
5.0/10 8.5 s
xai grok-imagine-image
Cost 0.02 $
Resolution 2752 x 1504
Time 8.5 s
Matania Judgment
Quality
8
Composition
7
Creativity
7
Text accuracy
2
Fidelity
4
Overall
5
Review
The model fails significantly on textual accuracy (text_accuracy) and fidelity: the characters do not match the requested writing systems, and the word 'PEACE' is not correctly rendered in the specified scripts. Although the visual aesthetics are clean, the image fails to meet the primary objective of translating the word into the five requested languages/scripts.
xai grok-imagine-image-pro
5.0/10 17.9 s
xai grok-imagine-image-pro
Cost 0.07 $
Resolution 2816 x 1536
Time 17.9 s
Matania Judgment
Quality
7
Composition
8
Creativity
7
Text accuracy
2
Fidelity
4
Overall
5
Review
The image fails significantly in terms of textual accuracy and prompt adherence. Although the aesthetics are polished, the model fails to coherently write the word 'PEACE' across the five requested scripts, mixing glyphs that do not correspond to the specified languages (notably for Arabic and Chinese). The multi-language calligraphy task is a major technical failure despite the overall high visual quality.

Mathematical formulas

image
google gemini-2.5-flash-image
6.5/10 5.9 s
google gemini-2.5-flash-image
Cost < 0.01 $
Resolution 1344 x 768
Time 5.9 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The overall aesthetics and composition are excellent, creating a realistic classroom atmosphere. However, prompt adherence is poor, as the specific equations requested (Maxwell, Schrödinger, Euler) are either absent or replaced by generic, illegible mathematical symbols. Textual precision is very low; the model prioritized the visual "mathematical style" at the expense of the accuracy of the requested formulas.
google gemini-3-pro-image-preview
6.5/10 39.4 s
google gemini-3-pro-image-preview
Cost < 0.01 $
Resolution 2752 x 1536
Time 39.4 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The image aesthetics are excellent, with a realistic rendering of the blackboard and chalk. However, prompt adherence is mediocre, as the specific requested equations (Maxwell, Schrödinger, Euler) are not clearly identifiable or are replaced by generic, illegible mathematical symbols. Textual precision is insufficient for a mathematical formula task, turning the equations into decorative text rather than exact scientific content.
google imagen-4.0-fast-generate-001
3.9 s
google imagen-4.0-fast-generate-001
Cost 0.02 $
Resolution 1408 x 768
Time 3.9 s
google imagen-4.0-generate-001
6.5/10 9.6 s
google imagen-4.0-generate-001
Cost 0.04 $
Resolution 1408 x 768
Time 9.6 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The image is aesthetically very successful, featuring beautiful light management and a "cinematic" composition. However, prompt adherence and textual precision are disappointing: the equations appear as abstract or illegible text, failing to respect the specific formulas requested (Maxwell, Schrödinger, Euler), which heavily penalizes the fidelity score.
google imagen-4.0-ultra-generate-001
6.5/10 12.5 s
google imagen-4.0-ultra-generate-001
Cost 0.08 $
Resolution 1408 x 768
Time 12.5 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The visual aesthetics are excellent, featuring a very realistic chalkboard rendering and a beautiful composition. However, textual precision is poor: although the atmosphere is well-captured, the equations are either partially incorrect or consist of mere incoherent mathematical symbols that fail to respect the rigor of the requested formulas (notably Maxwell or Schrödinger). Fidelity is penalized because the model prioritized artistic appeal at the expense of the explicitly requested scientific accuracy.
openai chatgpt-image-latest
6.5/10 43.3 s
openai chatgpt-image-latest
Cost 0.21 $
Resolution 1536 x 1024
Time 43.3 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The visual aesthetics are excellent and the composition is very successful, evoking a classic academic atmosphere. However, the textual precision is mediocre: while E=mc² is recognizable, the other equations (Maxwell, Schrödinger, Euler) are represented by generic and inconsistent mathematical symbols that do not respect the actual structure of these formulas. Fidelity is therefore impacted, as the model prioritized a "mathematical" visual appearance at the expense of the accuracy of the specific equations requested.
segmind ideogram-3
9.1/10 13.7 s
segmind ideogram-3
Cost 0.04 $
Resolution 1344 x 768
Time 13.7 s
Matania Judgment
Quality
9
Composition
9
Creativity
7
Text accuracy
8
Fidelity
10
Overall
9.13
Review
The image is of excellent technical quality, with a realistic rendering of the blackboard and the chalk. All requested equations are present and relatively well-spelled, perfectly adhering to the prompt's constraints. The composition is balanced and aesthetically very successful.
segmind seedream-4.5
4.9/10 19.1 s
segmind seedream-4.5
Cost 0.04 $
Resolution 2560 x 1472
Time 19.1 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.88
Review
The image is aesthetically pleasing with a beautiful chalkboard atmosphere, but it fails almost completely on a scientific level. The requested equations are replaced by incoherent mathematical text and nonsensical symbols (gibberish), resulting in very poor textual accuracy and prompt fidelity.
segmind seedream-v5-lite
4.9/10 35.0 s
segmind seedream-v5-lite
Cost 0.04 $
Resolution 2848 x 1600
Time 35.0 s
Matania Judgment
Quality
7
Composition
8
Creativity
6
Text accuracy
2
Fidelity
4
Overall
4.88
Review
The image is aesthetically pleasing with a nice blackboard atmosphere, but it fails almost entirely on textual accuracy. The equations are illegible or mathematically incorrect, which is critical for this prompt. Fidelity is low because, although the atmosphere is captured, the specific requested content (the precise formulas) is replaced by abstract and incoherent text.
xai grok-imagine-image
6.5/10 9.7 s
xai grok-imagine-image
Cost 0.02 $
Resolution 2752 x 1504
Time 9.7 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The overall aesthetics and image composition are excellent, creating a realistic academic atmosphere. However, textual accuracy is poor: although the mood is captured, the specific equations requested (Maxwell, Schrödinger, Euler) are replaced by incoherent mathematical text or abstract symbols. Fidelity is penalized because the model failed to reproduce the exact formulas requested in favor of a generic visual rendering.
xai grok-imagine-image-pro
6.5/10 17.4 s
xai grok-imagine-image-pro
Cost 0.07 $
Resolution 2816 x 1536
Time 17.4 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
6
Overall
6.5
Review
The image is aesthetically very successful, with a beautiful chalkboard atmosphere. However, prompt adherence is mediocre, as the requested equations are either poorly represented or replaced by generic, incoherent mathematical symbols. Textual precision is low, as the specific formulas (Maxwell, Schrödinger, Euler) are neither legible nor accurate.

Board game

image
google gemini-2.5-flash-image
8.9/10 7.0 s
google gemini-2.5-flash-image
Cost < 0.01 $
Resolution 1344 x 768
Time 7.0 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image perfectly respects all elements of the prompt (top-down view, hexagonal board, figurines, dice, and cards) with a very successful aesthetic. The technical quality and composition are excellent, creating an immersive atmosphere. The text_accuracy score reflects the illegibility of the inscriptions on the cards, which is common in image generation but necessary to note.
google gemini-3-pro-image-preview
8.9/10 35.0 s
google gemini-3-pro-image-preview
Cost < 0.01 $
Resolution 2752 x 1536
Time 35.0 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image is visually stunning, featuring excellent management of light and textures while perfectly adhering to all elements of the prompt (miniatures, dice, cards, hexagonal map). The top-down composition is impeccable. The text_accuracy score is reduced because the inscriptions on the cards are illegible scribbles, which is typical of current image models.
google imagen-4.0-fast-generate-001
4.4 s
google imagen-4.0-fast-generate-001
Cost 0.02 $
Resolution 1408 x 768
Time 4.4 s
google imagen-4.0-generate-001
9.5/10 8.7 s
google imagen-4.0-generate-001
Cost 0.04 $
Resolution 1408 x 768
Time 8.7 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
10
Fidelity
10
Overall
9.5
Review
The image perfectly adheres to all elements of the prompt: the top-down view, the figurines, the dice, the cards, and the hexagonal map are all present. The visual quality is exceptional, featuring remarkable sharpness and a balanced composition that truly evokes an ongoing board game session.
google imagen-4.0-ultra-generate-001
9.1/10 14.3 s
google imagen-4.0-ultra-generate-001
Cost 0.08 $
Resolution 1408 x 768
Time 14.3 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
7
Fidelity
10
Overall
9.13
Review
The image perfectly respects all elements of the prompt: the top-down view, the miniatures, the dice, the cards, and the hexagonal map are all present and coherent. The technical quality is excellent, featuring high sharpness and a very immersive board game aesthetic. The only minor drawback is the illegibility of the text on the cards, which is common in image generation, but it does not affect the overall composition.
openai chatgpt-image-latest
8.9/10 53.2 s
openai chatgpt-image-latest
Cost 0.22 $
Resolution 1536 x 1024
Time 53.2 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image perfectly respects all the elements requested in the prompt (top-down view, miniatures, dice, cards, and hexagonal map) with a highly successful aesthetic. The technical quality is excellent, although the text precision on the cards is poor, which is typical of image generation models. The composition is balanced and immersive for a fantasy board game.
segmind ideogram-3
9.1/10 13.1 s
segmind ideogram-3
Cost 0.04 $
Resolution 1344 x 768
Time 13.1 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
7
Fidelity
10
Overall
9.13
Review
The image perfectly adheres to all elements of the prompt: the top-down view, the miniatures, the dice, the cards, and the hexagonal map are all present and well-integrated. The technical quality is excellent, featuring great sharpness and a highly immersive board game aesthetic. The text on the cards is slightly blurry, which is normal for this type of generation, but the overall fidelity is exemplary.
segmind seedream-4.5
8.6/10 14.7 s
segmind seedream-4.5
Cost 0.04 $
Resolution 2560 x 1472
Time 14.7 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
5
Fidelity
10
Overall
8.63
Review
The image perfectly adheres to all elements of the prompt (top-down view, figurines, dice, cards, and hexagonal map) with a highly successful aesthetic. The visual quality and composition are excellent, offering immediate immersion. Only the text accuracy on the cards is poor, which is typical of image generation models, but this does not impact the overall fidelity to the requested concept.
segmind seedream-v5-lite
8.6/10 38.5 s
segmind seedream-v5-lite
Cost 0.04 $
Resolution 2848 x 1600
Time 38.5 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
5
Fidelity
10
Overall
8.63
Review
The image perfectly adheres to all elements of the prompt (top-down view, miniatures, dice, cards, and hexagonal map) with excellent fidelity. The composition is aesthetic and immersive, although the text accuracy on the cards is limited by the nature of image generation. The technical quality is overall very high and free of major artifacts.
xai grok-imagine-image
8.9/10 8.2 s
xai grok-imagine-image
Cost 0.02 $
Resolution 2752 x 1504
Time 8.2 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image perfectly respects all elements of the prompt (top-down view, figurines, dice, cards, and hexagonal map) with a highly successful aesthetic. The technical quality and composition are excellent, offering an immersive and detailed rendering. The text_accuracy score is reduced because the inscriptions on the cards and dice are unreadable abstract shapes, which is common in image generation.
xai grok-imagine-image-pro
8.9/10 15.0 s
xai grok-imagine-image-pro
Cost 0.07 $
Resolution 2816 x 1536
Time 15.0 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image perfectly adheres to all elements of the prompt (top-down view, miniatures, dice, cards, and hexagonal map) with excellent visual quality. The composition is balanced and immersive. The text_accuracy score is reduced because the inscriptions on the cards and dice are illegible visual artifacts, which is typical of image generation models.

Mobile interface

image
google gemini-2.5-flash-image
8.6/10 7.3 s
google gemini-2.5-flash-image
Cost < 0.01 $
Resolution 1344 x 768
Time 7.3 s
Matania Judgment
Quality
8
Composition
9
Creativity
8
Text accuracy
4
Fidelity
10
Overall
8.63
Review
The image adheres perfectly to the prompt, presenting three screens of a fitness application with a consistent dark theme. The composition and visual aesthetics are high-level, typical of a professional mockup. However, text accuracy is poor, with characters and numbers being illegible or inconsistent, which is common for image generation models.
google gemini-3-pro-image-preview
8.5/10 31.6 s
google gemini-3-pro-image-preview
Cost < 0.01 $
Resolution 2752 x 1536
Time 31.6 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
10
Overall
8.5
Review
The image adheres perfectly to the prompt, presenting three interface screens for a fitness application with a dark theme. The composition is aesthetic and professional, typical of a modern UI/UX mockup. However, text accuracy is low, featuring inconsistent characters and illegible filler text, which is common for image generation models.
google imagen-4.0-fast-generate-001
3.5 s
google imagen-4.0-fast-generate-001
Cost 0.02 $
Resolution 1408 x 768
Time 3.5 s
google imagen-4.0-generate-001
8.9/10 20.6 s
google imagen-4.0-generate-001
Cost 0.04 $
Resolution 1408 x 768
Time 20.6 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
5
Fidelity
10
Overall
8.88
Review
The image adheres perfectly to the prompt, presenting three fitness application screens with a consistent and aesthetic dark theme. The visual quality and composition are excellent, typical of a professional mockup. However, text accuracy is poor, featuring the generic and illegible text common to current image models, although this is frequent for this type of task.
google imagen-4.0-ultra-generate-001
9.0/10 27.8 s
google imagen-4.0-ultra-generate-001
Cost 0.08 $
Resolution 1408 x 768
Time 27.8 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
6
Fidelity
10
Overall
9
Review
The image adheres perfectly to the prompt, presenting three screens of a fitness application with an elegant dark theme. The visual quality and composition are excellent, offering a professional look. However, the text accuracy is imperfect, exhibiting characters typical of AI hallucinations, though this is common for this type of generation.
openai chatgpt-image-latest
8.8/10 47.4 s
openai chatgpt-image-latest
Cost 0.22 $
Resolution 1536 x 1024
Time 47.4 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
4
Fidelity
10
Overall
8.75
Review
The image adheres perfectly to the prompt by providing three distinct screens for a fitness application with a dark theme. The composition and visual quality are excellent, typical of a professional UI render. However, text accuracy is low, as the words are generic and illegible text forms, which is common for image generation models.
segmind ideogram-3
9.1/10 12.7 s
segmind ideogram-3
Cost 0.04 $
Resolution 1344 x 768
Time 12.7 s
Matania Judgment
Quality
9
Composition
9
Creativity
8
Text accuracy
7
Fidelity
10
Overall
9.13
Review
The image perfectly follows the prompt by presenting three distinct screens for a fitness application with a highly aesthetic dark theme. The visual quality is excellent and the composition is well-balanced, although the text accuracy shows some slight inaccuracies typical of image generation models.
segmind seedream-4.5
8.5/10 25.0 s
segmind seedream-4.5
Cost 0.04 $
Resolution 2560 x 1472
Time 25.0 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
10
Overall
8.5
Review
The image adheres perfectly to the prompt, presenting three screens of a fitness application with a highly aesthetic dark theme. The composition is balanced and professional, typical of a high-quality UI mockup. However, text precision is low, featuring illegible or inconsistent characters, which is a common issue in image generation models when rendering interfaces.
segmind seedream-v5-lite
8.5/10 37.8 s
segmind seedream-v5-lite
Cost 0.04 $
Resolution 2848 x 1600
Time 37.8 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
4
Fidelity
10
Overall
8.5
Review
The image adheres perfectly to the prompt, presenting three distinct screens with a dark theme and a consistent fitness aesthetic. The composition is excellent and highly professional, typical of a UI/UX mockup. However, text precision is poor, featuring illegible or incoherent characters, which is a common issue with image generation models.
xai grok-imagine-image
8.4/10 10.0 s
xai grok-imagine-image
Cost 0.02 $
Resolution 2752 x 1504
Time 10.0 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
3
Fidelity
10
Overall
8.38
Review
The image adheres perfectly to the prompt, presenting three fitness application screens with a highly aesthetic dark theme. The composition is excellent and professional, typical of a high-quality UI/UX mockup. However, text accuracy is very low, featuring illegible or inconsistent characters, which is common for image generation models.
xai grok-imagine-image-pro
8.4/10 20.8 s
xai grok-imagine-image-pro
Cost 0.07 $
Resolution 2816 x 1536
Time 20.8 s
Matania Judgment
Quality
8
Composition
9
Creativity
7
Text accuracy
3
Fidelity
10
Overall
8.38
Review
The image follows the prompt perfectly, presenting three screens of a fitness application with a highly aesthetic dark theme. The composition is professional and the visual quality is excellent, although the generated text is almost illegible and composed of incoherent characters, which is typical of current image generation models.
Code