Animations

Dynamic animations (Canvas/CSS) judged on smoothness and code quality.

5 tasks · 19 models tested · 95 results

Colored particles

html
anthropic claude-haiku-4-5-20251001
9.3/10 9.4 s
anthropic claude-haiku-4-5-20251001
Tokens 1 991
Source code 4.6 KB
Time 9.4 s
Matania Judgment
Animation
8
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.25
Review
The animation is fluid and adheres perfectly to the prompt's constraints, specifically the lack of mouse interaction and the absence of text. The visual rendering is aesthetic, featuring a beautiful use of neon colors against a dark background and a balanced composition. The code is clean, using exclusively CSS for the SVG animations, which is highly efficient.
anthropic claude-opus-4-6
9.4/10 40.5 s
anthropic claude-opus-4-6
Tokens 3 629
Source code 8.3 KB
Time 40.5 s
Matania Judgment
Animation
9
Rendering
9
Code quality
8
Creativity
9
Fidelity
10
Overall
9.38
Review
The animation is magnificent and perfectly follows the instructions: it is purely visual, contains no text, and completely ignores the mouse. The diversity of patterns (spirals, orbits, explosions) and the use of HSL color management create a rich and fluid visual spectacle. The code is well-structured; although truncated in the provided excerpt, it demonstrates robust particle simulation logic.
anthropic claude-opus-4-7
9.4/10 15.7 s
anthropic claude-opus-4-7
Tokens 1 880
Source code 2.4 KB
Time 15.7 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model perfectly adheres to all constraints, most notably the total absence of mouse interaction. The animation is fluid, complex, and aesthetically very successful, featuring a beautiful balance between the central core, the spiral arms, and the particles. The code is clean, optimized, and makes effective use of gradients and glow effects for a professional look.
anthropic claude-sonnet-4-6
9.6/10 39.4 s
anthropic claude-sonnet-4-6
Tokens 3 639
Source code 7.3 KB
Time 39.4 s
Matania Judgment
Animation
9
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.63
Review
The animation is absolutely stunning, far exceeding simple expectations for a "colored particles" prompt. The model perfectly adhered to the constraints of ignoring the mouse and excluding any text. The visual rendering is rich, featuring exemplary handling of gradients, glows, and depth, creating a hypnotic and fluid cosmic effect.
google gemini-flash-latest
9.4/10 16.8 s
google gemini-flash-latest
Tokens 1 091
Source code 3.9 KB
Time 16.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The animation is magnificent and scrupulously respects all prompt constraints, notably the absence of text and independence from the mouse. The visual rendering is high quality thanks to the use of SVG filters (glow, grain) that create an organic and elegant depth. The code is clean, utilizing Bézier curves to generate fluid and mathematically harmonious movements.
google gemini-flash-lite-latest
8.4/10 1.8 s
google gemini-flash-lite-latest
Tokens 543
Source code 1.7 KB
Time 1.8 s
Matania Judgment
Animation
7
Rendering
6
Code quality
9
Creativity
5
Fidelity
10
Overall
8.38
Review
The model perfectly adheres to the technical constraints: the animation is autonomous (no mouse interaction required) and contains no text. The code is very clean and lightweight (pure SVG), although the visual effect remains quite minimalist and geometric for a request for a "magnificent animation," lacking some of the complexity expected for "particles."
mistral mistral-large-latest
7.3/10 19.3 s
mistral mistral-large-latest
Tokens 972
Source code 2.8 KB
Time 19.3 s
Matania Judgment
Animation
6
Rendering
5
Code quality
3
Creativity
4
Fidelity
10
Overall
7.25
Review
The model perfectly respects the negative constraints (no text, no mouse interaction). However, the code quality is very poor: recreating and deleting the entire particle group every single frame (intensive DOM manipulation) is a technical aberration that is likely to cause stuttering and memory leaks. The visual rendering is basic, lacking the requested "magnificence."
mistral mistral-medium-latest
8.3/10 13.7 s
mistral mistral-medium-latest
Tokens 1 569
Source code 5.1 KB
Time 13.7 s
Matania Judgment
Animation
7
Rendering
8
Code quality
4
Creativity
7
Fidelity
10
Overall
8.25
Review
The model adheres perfectly to the prompt constraints, particularly the absence of text and the independence from mouse input. The visual effect is aesthetic, featuring a beautiful use of gradients and blur. However, the code quality is mediocre because the render function empties and reconstructs the entire SVG DOM on every frame (svg.innerHTML = ''), which is a very bad practice that causes major performance issues and potential memory leaks.
mistral mistral-small-latest
2.4/10 3.2 s
mistral mistral-small-latest
Tokens 605
Source code 1.3 KB
Time 3.2 s
Matania Judgment
Animation
4
Rendering
3
Code quality
6
Creativity
2
Fidelity
1
Overall
2.38
Review
The model failed completely on the main subject: it generated no particles, only a few simple geometric shapes (circles and a rectangle) with basic gradients. Although the constraint to ignore the mouse was respected, the result does not match the "colored particles" request at all.
openai gpt-4o-mini
2.9/10 11.7 s
openai gpt-4o-mini
Tokens 504
Source code 871 B
Time 11.7 s
Matania Judgment
Animation
4
Rendering
3
Code quality
6
Creativity
2
Fidelity
2
Overall
2.88
Review
The model completely failed to adhere to the requested subject ('Colored particles'), producing instead basic geometric shapes (circles, rectangles, triangles) that have nothing to do with a particle system. Although the technical constraint of not using the mouse was respected, the aesthetic quality is very low and the animation is rudimentary. Prompt fidelity is extremely poor, as the very essence of the request (particles) is absent.
openai gpt-5.4
9.0/10 44.2 s
openai gpt-5.4
Tokens 2 651
Source code 9.7 KB
Time 44.2 s
Matania Judgment
Animation
8
Rendering
9
Code quality
8
Creativity
7
Fidelity
10
Overall
9
Review
The model perfectly adhered to the negative constraints (no text, no mouse interaction). The use of SVG with blur filters and gradients creates a very aesthetic and soothing visual rendering, typical of a high-quality animated wallpaper. The code is clean and utilizes efficient native SVG animations.
openai gpt-5.4-mini
9.1/10 15.4 s
openai gpt-5.4-mini
Tokens 1 963
Source code 8.6 KB
Time 15.4 s
Matania Judgment
Animation
9
Rendering
9
Code quality
7
Creativity
8
Fidelity
10
Overall
9.13
Review
The animation is visually stunning and adheres perfectly to all prompt constraints, particularly the absence of text and the independence from mouse movement. The rendering is smooth, featuring excellent handling of gradients and blurs (glow effect). The code is of good quality, although it is incomplete at the end of the provided snippet (truncated), but the SVG implementation is clean and efficient.
openai gpt-5.4-nano
9.3/10 34.4 s
openai gpt-5.4-nano
Tokens 3 595
Source code 12.5 KB
Time 34.4 s
Matania Judgment
Animation
9
Rendering
9
Code quality
8
Creativity
8
Fidelity
10
Overall
9.25
Review
The animation is magnificent and strictly adheres to all constraints, specifically the lack of mouse interaction and the absence of text. The visual rendering is high quality, featuring sophisticated management of gradients and blurs to create an ethereal atmosphere. Although the code is truncated at the end, the logical structure and the implementation of the SVG filters are excellent.
openai gpt-5.4-pro
9.5/10 148.8 s
openai gpt-5.4-pro
Tokens 1 681
Source code 5.4 KB
Time 148.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
9
Fidelity
10
Overall
9.5
Review
The execution is perfect relative to the prompt: no mouse interaction, no text, and a purely visual aesthetic. The code demonstrates high technical maturity, utilizing flow fields and gradient compositions to achieve an organic and fluid rendering. The depth effect created by the stars and colored orbs is magnificent.
openai gpt-5.5
9.8/10 44.4 s
openai gpt-5.5
Tokens 1 703
Source code 5.5 KB
Time 44.4 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.75
Review
The animation is absolutely sublime, featuring masterful management of layers, bloom effects, and colors that create a hypnotic depth. The code is extremely clean, utilizing complex mathematical functions (sinusoids, Bézier curves) to generate organic and fluid movement. All prompt constraints were respected: the animation is autonomous, contains no text, and does not react to the mouse.
openai gpt-5.5-pro
9.5/10 147.0 s
openai gpt-5.5-pro
Tokens 2 248
Source code 7.6 KB
Time 147.0 s
Matania Judgment
Animation
9
Rendering
10
Code quality
8
Creativity
9
Fidelity
10
Overall
9.5
Review
The animation is exceptionally beautiful, featuring complex layer management (stars, ribbons, bloom, particles) that creates an immersive visual depth. Constraint compliance is perfect: there is no mouse interaction, and the aesthetic is purely visual. The code is well-structured, despite the truncation at the end of the provided snippet.
productivia matania-latest
8.3/10 9.0 s
productivia matania-latest
Tokens 1 260
Source code 4.2 KB
Time 9.0 s
Matania Judgment
Animation
7
Rendering
5
Code quality
8
Creativity
6
Fidelity
10
Overall
8.25
Review
The code is technically very clean, intelligently utilizing a blend of SMIL and JavaScript to generate dynamic SVG elements without memory leaks. Prompt adherence is perfect, respecting the absence of text and the constraint of no mouse interaction. However, in the absence of a screenshot, the rendering score is limited by default, even though the code suggests a consistent and elegant spatial aesthetic.
xai grok-4-1-fast-non-reasoning
8.8/10 23.6 s
xai grok-4-1-fast-non-reasoning
Tokens 1 956
Source code 6.6 KB
Time 23.6 s
Matania Judgment
Animation
8
Rendering
9
Code quality
6
Creativity
7
Fidelity
10
Overall
8.75
Review
The animation is visually very successful, fluid, and perfectly adheres to the constraint of not using the mouse. However, the source code is truncated at the end (incomplete code), which prevents a perfect technical rendering and affects the code quality score.
xai grok-4-1-fast-reasoning
8.5/10 91.4 s
xai grok-4-1-fast-reasoning
Tokens 3 586
Source code 13.0 KB
Time 91.4 s
Matania Judgment
Animation
7
Rendering
8
Code quality
6
Creativity
7
Fidelity
10
Overall
8.5
Review
The model perfectly adhered to the negative constraints (no text, no mouse interaction). The animation is aesthetic and soothing, although the code is incomplete (truncated at the end), which prevents a full evaluation of the structure. The visual rendering via SVG is clean, and the glow effects are well-managed.

Analog clock

html
anthropic claude-haiku-4-5-20251001
9.4/10 5.5 s
anthropic claude-haiku-4-5-20251001
Tokens 1 420
Source code 2.4 KB
Time 5.5 s
Matania Judgment
Animation
10
Rendering
8
Code quality
10
Creativity
7
Fidelity
10
Overall
9.38
Review
The model scrupulously adheres to all technical constraints, notably the use of ES5, the exact code structure, and the smooth animation of seconds through the integration of milliseconds. The visual rendering is clean and futuristic with effective neon colors, although the aesthetic remains fairly classic for Canvas. Prompt fidelity is perfect.
anthropic claude-opus-4-6
9.9/10 25.4 s
anthropic claude-opus-4-6
Tokens 2 577
Source code 5.0 KB
Time 25.4 s
Matania Judgment
Animation
10
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.88
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, the strict output format, and the smooth animation of the second. The visual rendering is exceptional, featuring a coherent futuristic aesthetic (neons, gradients, drop shadows) and precise fluidity management via milliseconds.
anthropic claude-opus-4-7
9.6/10 18.4 s
anthropic claude-opus-4-7
Tokens 2 477
Source code 3.1 KB
Time 18.4 s
Matania Judgment
Animation
10
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.63
Review
The model scrupulously adheres to all technical constraints, notably the use of ES5, the exact HTML structure, and the fluid movement of the second hand thanks to the integration of milliseconds. The visual rendering is aesthetic and perfectly matches the 'futuristic' theme, with well-managed glow effects. The code is clean, optimized, and free of memory leaks.
anthropic claude-sonnet-4-6
9.9/10 20.9 s
anthropic claude-sonnet-4-6
Tokens 2 700
Source code 5.1 KB
Time 20.9 s
Matania Judgment
Animation
10
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.88
Review
The model perfectly adhered to all technical constraints, notably the use of ES5, the canvas structure, and the smooth second-by-second animation achieved through millisecond integration. The visual rendering is exceptional, featuring a very consistent neon/futuristic style and excellent management of gradients and glow effects.
google gemini-flash-latest
9.6/10 15.3 s
google gemini-flash-latest
Tokens 1 263
Source code 2.0 KB
Time 15.3 s
Matania Judgment
Animation
10
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.63
Review
The model strictly adheres to all technical constraints, particularly the use of ES5 and the strict output format. The animation of the second hand is perfectly smooth thanks to the integration of milliseconds into the angle calculation. The visual rendering is aesthetic, featuring a successful neon/futuristic effect that aligns well with the request.
google gemini-flash-lite-latest
9.1/10 3.1 s
google gemini-flash-lite-latest
Tokens 893
Source code 1.2 KB
Time 3.1 s
Matania Judgment
Animation
10
Rendering
8
Code quality
9
Creativity
6
Fidelity
10
Overall
9.13
Review
The model strictly adheres to all technical constraints, notably the use of ES5 JavaScript, the strict output format, and the smooth animation of the second hand achieved through millisecond integration. The visual rendering is clean and futuristic with its cyan/black contrast, although the absence of numbers or text on the dial slightly limits the requested creativity.
mistral mistral-large-latest
9.4/10 12.2 s
mistral mistral-large-latest
Tokens 1 071
Source code 2.8 KB
Time 12.2 s
Matania Judgment
Animation
10
Rendering
8
Code quality
10
Creativity
7
Fidelity
10
Overall
9.38
Review
The model strictly adheres to all technical constraints, notably the use of ES5, the raw output structure, and the smooth animation of the second hand achieved through millisecond integration. The visual rendering is clean, and the futuristic aesthetic is well-conveyed by the contrast of neon colors against a dark background, although the composition remains fairly classic.
mistral mistral-medium-latest
9.6/10 6.3 s
mistral mistral-medium-latest
Tokens 984
Source code 2.4 KB
Time 6.3 s
Matania Judgment
Animation
10
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5 JavaScript, the exact HTML structure, and the smooth animation of the second hand through the integration of milliseconds. The visual rendering is highly successful, featuring a consistent neon/futuristic aesthetic and excellent readability. The code is clean, well-structured, and makes efficient use of Canvas rotation functions.
mistral mistral-small-latest
9.3/10 5.1 s
mistral mistral-small-latest
Tokens 1 000
Source code 2.5 KB
Time 5.1 s
Matania Judgment
Animation
10
Rendering
8
Code quality
9
Creativity
7
Fidelity
10
Overall
9.25
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, the exact HTML structure, and the smooth animation of the seconds through the integration of milliseconds. The rendering is clean and aesthetic, although the 'futuristic' style remains quite sober and classic. The code quality is excellent, featuring an efficient animation loop.
openai gpt-4o-mini
3.1/10 9.9 s
openai gpt-4o-mini
Tokens 687
Source code 1.3 KB
Time 9.9 s
Matania Judgment
Animation
2
Rendering
4
Code quality
5
Creativity
2
Fidelity
3
Overall
3.13
Review
The model failed on the major constraint of smooth animation (point 4): the use of setInterval(..., 1000) results in a jerky, rather than continuous, second. The style is extremely basic and fails to respect the requested "futuristic" look. Finally, although the code is in ES5, it does not meet the visual expectations for a project of this type.
openai gpt-5.4
2.6/10 25.3 s
openai gpt-5.4
Tokens 2 046
Source code 6.6 KB
Time 25.3 s
Matania Judgment
Animation
0
Rendering
10
Code quality
2
Creativity
9
Fidelity
0
Overall
2.63
Review
The model failed critically on faithfulness: the code is incomplete, cutting off abruptly in the middle of the drawHUD function without ever calling requestAnimationFrame or closing the tags, which renders the clock completely inoperative. While the visual design (rendering) and creativity are excellent, the absence of an animation loop and time-update logic makes the result a failure regarding the technical constraints.
openai gpt-5.4-mini
9.5/10 5.8 s
openai gpt-5.4-mini
Tokens 1 043
Source code 2.7 KB
Time 5.8 s
Matania Judgment
Animation
10
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, the strict output format, and the smooth animation of the second hand achieved through millisecond integration. The visual rendering is high-quality, featuring a very successful neon/futuristic aesthetic, and the code is optimized for Canvas.
openai gpt-5.4-nano
9.6/10 22.7 s
openai gpt-5.4-nano
Tokens 2 785
Source code 9.5 KB
Time 22.7 s
Matania Judgment
Animation
10
Rendering
9
Code quality
9
Creativity
9
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, the strict output format, and the smooth animation of the second hand achieved through millisecond integration. The visual rendering is exceptional, featuring a highly polished neon/futuristic style and excellent management of gradients and glow effects.
openai gpt-5.4-pro
5.9/10 372.7 s
openai gpt-5.4-pro
Tokens 1 872
Source code 5.9 KB
Time 372.7 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
10
Fidelity
2
Overall
5.88
Review
The model produced an absolutely magnificent visual rendering and extremely smooth animation (the second hand moves continuously thanks to the milliseconds). However, the code is truncated at the end (syntax error: 'x.' followed by nothing), which prevents actual execution and constitutes a critical failure of the completeness constraint. The fidelity score is severely impacted by this incomplete code.
openai gpt-5.5
9.8/10 21.1 s
openai gpt-5.5
Tokens 1 102
Source code 2.9 KB
Time 21.1 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, including the strict output format, the use of ES5, and the smooth animation of the second hand via milliseconds. The visual rendering is exceptional, featuring a highly successful neon/futuristic style and a balanced composition. The code is compact and efficient, making judicious use of glow effects (shadowBlur) to achieve the requested aesthetic.
openai gpt-5.5-pro
9.9/10 264.7 s
openai gpt-5.5-pro
Tokens 1 660
Source code 5.1 KB
Time 264.7 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
10
Fidelity
10
Overall
9.88
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, raw HTML structure, and the smooth second-hand animation. The visual rendering is exceptional, exceeding expectations with a neon effect, animated stars, and a highly consistent futuristic aesthetic. The handling of the current time and Roman numerals is impeccable.
productivia matania-latest
8.9/10 7.5 s
productivia matania-latest
Tokens 1 107
Source code 2.9 KB
Time 7.5 s
Matania Judgment
Animation
10
Rendering
5
Code quality
9
Creativity
7
Fidelity
10
Overall
8.88
Review
The code is excellent and strictly adheres to all technical constraints (ES5, exact structure, smooth second-by-second animation via milliseconds). The 'rendering' score is penalized because no screenshot was provided to validate the final aesthetics, even though the code suggests a consistent futuristic look.
xai grok-4-1-fast-non-reasoning
9.8/10 12.8 s
xai grok-4-1-fast-non-reasoning
Tokens 1 611
Source code 4.9 KB
Time 12.8 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, raw HTML structure, and the smooth animation of the second hand achieved through millisecond integration. The visual rendering is exceptional, featuring a well-executed neon/futuristic style via gradients and Canvas glow effects (shadowBlur). The code is clean and efficient.
xai grok-4-1-fast-reasoning
9.6/10 40.8 s
xai grok-4-1-fast-reasoning
Tokens 1 282
Source code 3.6 KB
Time 40.8 s
Matania Judgment
Animation
10
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5, the exact code structure, and the smooth second-by-second animation via milliseconds. The visual output is aesthetic, respecting the futuristic theme with successful glow effects on Canvas. The code is clean, optimized with requestAnimationFrame, and shows no apparent memory leaks.

Sine wave

html
anthropic claude-haiku-4-5-20251001
6.5/10 4.4 s
anthropic claude-haiku-4-5-20251001
Tokens 973
Source code 1.5 KB
Time 4.4 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
4
Overall
6.5
Review
The code is technically excellent, fluid, and aesthetically very successful with a beautiful glow effect. However, it fails on the prompt's primary constraint: the mouseover interaction does not influence the wave itself (the wave scrolls autonomously), with the mouse only creating a passive decorative halo. Furthermore, the prompt requested a mouseover animation, which implies a reaction from the wave to the mouse—something that has not been implemented here.
anthropic claude-opus-4-6
9.5/10 23.1 s
anthropic claude-opus-4-6
Tokens 2 238
Source code 4.5 KB
Time 23.1 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the requested HTML structure. The mouse-over effect is highly sophisticated, creating a dynamic interaction that amplifies the waves and adds a radial glow. The code is clean and well-structured, and the aesthetics are high-quality, featuring a gradient background and subtle particles.
anthropic claude-opus-4-7
9.4/10 10.6 s
anthropic claude-opus-4-7
Tokens 1 398
Source code 1.4 KB
Time 10.6 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model strictly adheres to all technical constraints, particularly the use of ES5 and the single output format. The implementation of the hover interaction is excellent, featuring dynamic amplitude modulation and an organic jitter effect that goes beyond the initial request. The visual rendering is highly aesthetic thanks to the use of a trailing effect (alpha on the background) and glows (shadowBlur).
anthropic claude-sonnet-4-6
9.6/10 16.8 s
anthropic claude-sonnet-4-6
Tokens 1 983
Source code 3.6 KB
Time 16.8 s
Matania Judgment
Animation
9
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the requested canvas structure. The mouse hover interaction effect is highly sophisticated, adding a dynamic deformation to the waves that exceeds the simple expectations of the prompt. The visual rendering is aesthetically superb, featuring glow effects and a well-crafted gradient background.
google gemini-flash-latest
9.5/10 18.6 s
google gemini-flash-latest
Tokens 899
Source code 1.4 KB
Time 18.6 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model strictly adheres to all technical constraints, specifically the use of ES5 and the requested HTML structure. The 'mouse over' implementation is highly successful, dynamically influencing both amplitude and frequency, which goes beyond the basic requirements. The animation is fluid, and the aesthetics are polished, featuring a beautiful motion blur effect achieved via transparent fillRect.
google gemini-flash-lite-latest
6.0/10 2.2 s
google gemini-flash-lite-latest
Tokens 578
Source code 540 B
Time 2.2 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
6
Fidelity
4
Overall
6
Review
The code is clean and adheres to the technical constraints (ES5, structure, canvas), but it fails on the major interaction requirement: the wave does not react to 'mouse over' at all as requested. While the aesthetics are fine, the absence of the primary interactive functionality drastically lowers the fidelity score.
mistral mistral-large-latest
6.1/10 4.8 s
mistral mistral-large-latest
Tokens 548
Source code 875 B
Time 4.8 s
Matania Judgment
Animation
9
Rendering
8
Code quality
10
Creativity
6
Fidelity
4
Overall
6.13
Review
The code is clean, well-structured in ES5, and the animation is smooth. However, the model completely ignored the prompt's major constraint: the 'mouse over' interaction. The wave scrolls autonomously without reacting to the mouse, resulting in a very low fidelity score despite the technical quality of the rest.
mistral mistral-medium-latest
9.1/10 4.8 s
mistral mistral-medium-latest
Tokens 699
Source code 1.4 KB
Time 4.8 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
7
Fidelity
10
Overall
9.13
Review
The model strictly adheres to all technical constraints, specifically the use of ES5, raw HTML structure, and mouseover interaction. The animation is smooth, and the wave overlay effect is aesthetically successful. The addition of new random waves upon hovering adds a relevant interactive dimension that goes beyond the basic requirements.
mistral mistral-small-latest
5.5/10 2.6 s
mistral mistral-small-latest
Tokens 510
Source code 713 B
Time 2.6 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
6
Fidelity
3
Overall
5.5
Review
The model fails on the prompt's primary constraint: the animation does not trigger on 'mouse over' but is automatic instead. Although the code is clean, adheres to ES5, and provides a smooth and aesthetic visual rendering, the absence of the requested interaction constitutes a critical fidelity flaw.
openai gpt-4o-mini
4.9/10 3.9 s
openai gpt-4o-mini
Tokens 434
Source code 410 B
Time 3.9 s
Matania Judgment
Animation
8
Rendering
6
Code quality
9
Creativity
4
Fidelity
3
Overall
4.88
Review
The model fails on the prompt's major constraint: the animation does not react to the 'mouse over' at all. Although the code is clean, fluid, and adheres to both ES5 syntax and the requested format, the absence of the primary user interaction drastically lowers the fidelity score.
openai gpt-5.4
9.9/10 11.7 s
openai gpt-5.4
Tokens 1 014
Source code 2.7 KB
Time 11.7 s
Matania Judgment
Animation
10
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.88
Review
The model perfectly adhered to all technical constraints, including the use of ES5 and the specified output format. The mouse-over effect is elegantly implemented with smooth easing that affects both the amplitude and the brightness of the waves. The visual rendering is professional, featuring a harmonious management of colors, gradients, and glow effects.
openai gpt-5.4-mini
9.4/10 3.8 s
openai gpt-5.4-mini
Tokens 686
Source code 1.4 KB
Time 3.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model strictly adheres to all technical constraints, specifically the use of ES5 and the requested raw HTML structure. The 'mouse over' effect is subtly integrated through amplitude variation, and the aesthetics are highly successful, featuring gradients and glow effects that exceed simple expectations. The animation is perfectly fluid and the code is optimized.
openai gpt-5.4-nano
9.6/10 17.4 s
openai gpt-5.4-nano
Tokens 2 035
Source code 6.7 KB
Time 17.4 s
Matania Judgment
Animation
9
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all constraints, including the use of ES5 JavaScript and the specific structure. The visual rendering is exceptional, featuring sophisticated handling of light (glow), motion, and mouse interaction. The animation is fluid, and the aesthetic exceeds expectations for a simple sine wave.
openai gpt-5.4-pro
9.9/10 279.9 s
openai gpt-5.4-pro
Tokens 1 625
Source code 5.1 KB
Time 279.9 s
Matania Judgment
Animation
10
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.88
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the requested HTML structure. The implementation of the mouse hover interaction is exceptional, featuring smooth easing that affects the amplitude, speed, and glow of the waves. The visual rendering is professional-grade, with sophisticated management of glow effects and composition.
openai gpt-5.5
9.8/10 16.9 s
openai gpt-5.5
Tokens 1 093
Source code 3.0 KB
Time 16.9 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the requested structure. The mouseover interaction is extremely fluid and creative, adding an elegant distortion dimension to the wave. The visual rendering is high-quality, featuring sophisticated management of light effects (glow) and composition.
openai gpt-5.5-pro
9.9/10 181.0 s
openai gpt-5.5-pro
Tokens 1 121
Source code 3.1 KB
Time 181.0 s
Matania Judgment
Animation
10
Rendering
10
Code quality
9
Creativity
10
Fidelity
10
Overall
9.88
Review
The model has far exceeded expectations. The animation is exemplarily smooth, featuring a parallax effect and highly sophisticated hover interaction (mouse influence on the wave shape). The code strictly adheres to all technical constraints (ES5, 400x400 canvas, requestAnimationFrame), and the aesthetic is worthy of a professional-grade component.
productivia matania-latest
9.0/10 5.3 s
productivia matania-latest
Tokens 837
Source code 2.0 KB
Time 5.3 s
Matania Judgment
Animation
9
Rendering
6
Code quality
9
Creativity
8
Fidelity
10
Overall
9
Review
The code is excellent and scrupulously adheres to all technical constraints, including the use of ES5 and the requested structure. The hover interaction is well-implemented, featuring a highlight effect and a light gradient. The rendering score was penalized because no screenshot was provided to validate the final aesthetics.
xai grok-4-1-fast-non-reasoning
9.4/10 4.6 s
xai grok-4-1-fast-non-reasoning
Tokens 622
Source code 1.1 KB
Time 4.6 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model scrupulously adheres to all technical constraints, notably the use of ES5 JavaScript and the strict output format. The animation is fluid, and the wave effect reacts elegantly to mouse movement, creating a satisfying visual interaction. The aesthetics are polished, featuring good use of HSL colors and glow effects (shadowBlur).
xai grok-4-1-fast-reasoning
9.5/10 46.2 s
xai grok-4-1-fast-reasoning
Tokens 790
Source code 1.8 KB
Time 46.2 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
9
Fidelity
10
Overall
9.5
Review
The model strictly adheres to all technical constraints, including the raw HTML format and the use of ES5 JavaScript. The mouse-over ripple effect is an excellent creative interpretation of the prompt, going beyond simple basic animation. The visual rendering is aesthetic, featuring very successful gradients and glow effects.

Starry galaxy

html
anthropic claude-haiku-4-5-20251001
9.1/10 3.0 s
anthropic claude-haiku-4-5-20251001
Tokens 862
Source code 1.2 KB
Time 3.0 s
Matania Judgment
Animation
9
Rendering
8
Code quality
10
Creativity
6
Fidelity
10
Overall
9.13
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5 JavaScript, the requested HTML structure, and the exact number of stars. The animation is smooth, and the visual rendering accurately follows the requested aesthetic (black background, bluish/white stars). The code is clean, optimized, and free of any memory leaks.
anthropic claude-opus-4-6
9.5/10 17.0 s
anthropic claude-opus-4-6
Tokens 1 637
Source code 3.1 KB
Time 17.0 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints (ES5, 400x400 canvas, 200 stars, raw HTML output). The implementation of the depth effect via size and twinkling, as well as the management of bluish hues, exceeds the simple expectations of the prompt. The code is clean and optimized, and the animation is smooth with a very aesthetic visual rendering.
anthropic claude-opus-4-7
9.1/10 6.4 s
anthropic claude-opus-4-7
Tokens 996
Source code 721 B
Time 6.4 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
7
Fidelity
10
Overall
9.13
Review
The model strictly adheres to all technical constraints: the star count, output format, use of ES5, and canvas structure. The depth effect is well-managed through variations in size and color (bluish tones for distant stars), and the animation is smooth thanks to differential rotation based on distance. The code is concise, efficient, and perfectly functional.
anthropic claude-sonnet-4-6
9.5/10 11.1 s
anthropic claude-sonnet-4-6
Tokens 1 217
Source code 1.9 KB
Time 11.1 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the raw output format. The visual effect is excellent, featuring intelligent depth management through size and brightness, as well as an elegant rendering of the galactic spiral thanks to the addition of a central gradient and a halo effect.
google gemini-flash-latest
9.5/10 12.7 s
google gemini-flash-latest
Tokens 770
Source code 2.2 KB
Time 12.7 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model strictly adheres to all technical constraints, notably the use of ES5 JavaScript, raw HTML format, and the exact number of stars. The elliptical perspective effect and the central gradient add an aesthetic dimension that goes beyond the simple request, creating a visually compelling and fluid galaxy.
google gemini-flash-lite-latest
9.0/10 2.9 s
google gemini-flash-lite-latest
Tokens 738
Source code 870 B
Time 2.9 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
6
Fidelity
10
Overall
9
Review
The model perfectly adhered to all technical constraints, specifically the use of JavaScript ES5 and the single output format. The animation is fluid, and the differential speed effect based on distance from the center effectively simulates a galactic rotation. The visual rendering is clean, albeit simple, faithfully respecting the requested color palette.
mistral mistral-large-latest
9.3/10 7.1 s
mistral mistral-large-latest
Tokens 689
Source code 1.4 KB
Time 7.1 s
Matania Judgment
Animation
9
Rendering
8
Code quality
10
Creativity
7
Fidelity
10
Overall
9.25
Review
The model scrupulously adheres to all technical constraints: ES5 JavaScript, a 400x400 canvas, 200 stars, and raw output format. The implementation of radial gradients for the shimmer effect is excellent, providing great visual depth despite the code's simplicity. The animation is smooth, and the code structure is perfectly clean.
mistral mistral-medium-latest
6.4/10 4.6 s
mistral mistral-medium-latest
Tokens 694
Source code 1.4 KB
Time 4.6 s
Matania Judgment
Animation
6
Rendering
7
Code quality
9
Creativity
5
Fidelity
6
Overall
6.38
Review
The model respects the structure and technical constraints (ES5, canvas, 200 stars), but fails on the mathematical logic of the animation: by multiplying the angle by time within the cos/sin functions, it creates exponential acceleration instead of a slow, constant rotation. The visual rendering is correct with the requested colors, but the movement quickly becomes chaotic and unpredictable.
mistral mistral-small-latest
8.3/10 2.4 s
mistral mistral-small-latest
Tokens 542
Source code 798 B
Time 2.4 s
Matania Judgment
Animation
7
Rendering
7
Code quality
7
Creativity
5
Fidelity
10
Overall
8.25
Review
The model strictly adheres to all technical constraints: raw HTML format, 400x400 canvas, ES5 JavaScript, and the 200 stars. The animation is functional but uses setInterval instead of requestAnimationFrame as requested, which is a slight deviation from rule 5. The visual rendering is compliant (white/bluish stars on a black background) but remains very basic.
openai gpt-4o-mini
7.9/10 5.9 s
openai gpt-4o-mini
Tokens 537
Source code 883 B
Time 5.9 s
Matania Judgment
Animation
6
Rendering
5
Code quality
8
Creativity
4
Fidelity
10
Overall
7.88
Review
The model strictly adheres to all technical constraints (ES5, 400x400 canvas, 200 stars, raw HTML structure). However, implementing the shimmer via Math.random() on every frame creates an extremely aggressive and visually unpleasant flickering (noise/stroboscopic effect) rather than an elegant twinkling effect. The rendering is functional but lacks true visual depth despite the variation in size.
openai gpt-5.4
9.5/10 11.0 s
openai gpt-5.4
Tokens 908
Source code 2.5 KB
Time 11.0 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5 and raw HTML structure. The visual rendering is excellent, achieving a successful sense of depth through size variation and the addition of a galactic core that exceeds the expectations of the initial prompt. The animation is smooth, and the twinkle effect adds a very pleasant touch of realism.
openai gpt-5.4-mini
9.1/10 3.7 s
openai gpt-5.4-mini
Tokens 607
Source code 1.0 KB
Time 3.7 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
7
Fidelity
10
Overall
9.13
Review
The model strictly adheres to all technical constraints, including the use of ES5 JavaScript, raw HTML format, and the specific star count. The depth effect is well-managed through size variation and the addition of light streaks for larger stars. The animation is smooth, and the aesthetic perfectly respects the requested color palette.
openai gpt-5.4-nano
9.8/10 8.1 s
openai gpt-5.4-nano
Tokens 1 034
Source code 5.6 KB
Time 8.1 s
Matania Judgment
Animation
9
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, including the use of ES5 JavaScript and the specific canvas structure. The visual rendering is exceptional for a simple script, featuring sophisticated handling of depth, twinkling, and halos that far exceeds basic expectations. The animation is fluid, and the aesthetic composition is worthy of a true galaxy simulation.
openai gpt-5.4-pro
9.5/10 378.7 s
openai gpt-5.4-pro
Tokens 1 021
Source code 2.6 KB
Time 378.7 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the use of ES5 and the requested structure. The implementation of rotation, depth via scaling, and the shimmering effect is of high mathematical quality. The visual rendering is elegant, featuring a subtle galactic core and very clean management of the bluish tones.
openai gpt-5.5
9.4/10 12.0 s
openai gpt-5.5
Tokens 615
Source code 1.1 KB
Time 12.0 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model strictly adheres to all technical constraints (ES5, 400x400 canvas, 200 stars, raw HTML only). The mathematical implementation is excellent, utilizing spiral arms to simulate a spiral galaxy rather than a simple circle, which exceeds creative expectations. The animation is smooth, and the variations in size and color provide beautiful visual depth.
openai gpt-5.5-pro
9.4/10 231.8 s
openai gpt-5.5-pro
Tokens 719
Source code 1.5 KB
Time 231.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model perfectly adhered to all technical constraints, notably the use of ES5 JavaScript and the raw output format. The implementation of the depth effect through variations in size and brightness is excellent, and the animation is smooth thanks to intelligent time-delta management. The visual rendering is aesthetic, featuring a beautiful blend of bluish and white hues against a deep black background.
productivia matania-latest
9.0/10 4.4 s
productivia matania-latest
Tokens 720
Source code 1.5 KB
Time 4.4 s
Matania Judgment
Animation
9
Rendering
6
Code quality
10
Creativity
7
Fidelity
10
Overall
9
Review
The code is excellent and strictly adheres to all technical constraints (ES5, format, star count, colors). The animation will be smooth thanks to requestAnimationFrame. The rendering score is penalized because the absence of a screenshot makes it impossible to validate the actual aesthetics, even though the code suggests a clean and compliant output.
xai grok-4-1-fast-non-reasoning
9.4/10 3.8 s
xai grok-4-1-fast-non-reasoning
Tokens 612
Source code 2.1 KB
Time 3.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
9
Creativity
8
Fidelity
10
Overall
9.38
Review
The model perfectly adhered to all technical constraints, including the output format, star count, use of ES5, and the 400x400 canvas. The implementation of the logarithmic spiral and radial gradients exceeds expectations for such a simple prompt, delivering a highly aesthetic visual result and smooth animation.
xai grok-4-1-fast-reasoning
9.3/10 34.6 s
xai grok-4-1-fast-reasoning
Tokens 642
Source code 1.2 KB
Time 34.6 s
Matania Judgment
Animation
9
Rendering
8
Code quality
9
Creativity
8
Fidelity
10
Overall
9.25
Review
The model perfectly adhered to all technical constraints, including the use of ES5 and the strict output format. The implementation of the spiral effect (spiralTwist) and depth via size/brightness exceeds the prompt's minimum expectations. The animation is smooth and the visual rendering is aesthetic, although the stars are slightly large relative to the canvas size.

Animated CSS loader

html
anthropic claude-haiku-4-5-20251001
9.5/10 13.7 s
anthropic claude-haiku-4-5-20251001
Tokens 1 382
Source code 2.3 KB
Time 13.7 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the prohibition of SVG/JS and the strict output format. The aesthetics are highly successful, featuring smooth layered animations and a modern background gradient. The code is clean and well-structured, resulting in a visual effect that is both elegant and professional.
anthropic claude-opus-4-6
6.6/10 19.7 s
anthropic claude-opus-4-6
Tokens 2 198
Source code 3.7 KB
Time 19.7 s
Matania Judgment
Animation
9
Rendering
10
Code quality
9
Creativity
9
Fidelity
4
Overall
6.63
Review
The visual rendering and animation are exceptional, featuring a modern aesthetic and highly successful depth effects. However, the model failed on a critical fidelity constraint: the code is truncated at the end (it cuts off abruptly at the first div of the ring), which renders the component incomplete and non-functional in its entirety.
anthropic claude-opus-4-7
9.5/10 7.1 s
anthropic claude-opus-4-7
Tokens 1 199
Source code 1.1 KB
Time 7.1 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the prohibition of SVG and JavaScript. The aesthetics are very modern, featuring successful gradients and glow effects. The animation is fluid and complex thanks to the use of pseudo-elements and opposing directions, creating a rich visual effect.
anthropic claude-sonnet-4-6
9.8/10 13.8 s
anthropic claude-sonnet-4-6
Tokens 1 885
Source code 3.4 KB
Time 13.8 s
Matania Judgment
Animation
9
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, specifically the prohibition of SVG and JavaScript. The rendering is visually exceptional, featuring masterful handling of gradients, drop shadows (glow effects), and depth. The animation is fluid, complex, and highly elegant.
google gemini-flash-latest
9.5/10 10.2 s
google gemini-flash-latest
Tokens 1 077
Source code 1.6 KB
Time 10.2 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the absence of SVG/JS and the strict output format. The aesthetics are modern, featuring excellent color usage and smooth layered animations. The code is clean and optimized, and the visual rendering is highly professional.
google gemini-flash-lite-latest
8.5/10 2.1 s
google gemini-flash-lite-latest
Tokens 621
Source code 552 B
Time 2.1 s
Matania Judgment
Animation
8
Rendering
7
Code quality
9
Creativity
4
Fidelity
10
Overall
8.5
Review
The model strictly adheres to all technical constraints, notably the absence of SVG/JS and the strict output format. The animation is smooth and clean, although the design remains very conventional and lacks originality (low creativity). The visual rendering is compliant and elegant, respecting the requested dimensions.
mistral mistral-large-latest
8.8/10 4.2 s
mistral mistral-large-latest
Tokens 548
Source code 721 B
Time 4.2 s
Matania Judgment
Animation
8
Rendering
7
Code quality
9
Creativity
6
Fidelity
10
Overall
8.75
Review
The model strictly adheres to all technical constraints, notably the absence of SVG and the exclusive use of HTML/CSS. The pulsing animation is smooth and elegant, even though the design remains quite minimalist. The code is clean, well-structured, and perfectly centered as requested.
mistral mistral-medium-latest
8.8/10 3.4 s
mistral mistral-medium-latest
Tokens 647
Source code 1.1 KB
Time 3.4 s
Matania Judgment
Animation
7
Rendering
8
Code quality
9
Creativity
6
Fidelity
10
Overall
8.75
Review
The model strictly adheres to all technical constraints, notably the absence of SVG/JS and the strict output format. The animation combines rotation and pulsation smoothly, although the visual effect remains fairly conventional. The code is clean and well-structured, and the rendering is aesthetically modern, meeting all expectations.
mistral mistral-small-latest
4.5/10 2.5 s
mistral mistral-small-latest
Tokens 573
Source code 821 B
Time 2.5 s
Matania Judgment
Animation
8
Rendering
7
Code quality
2
Creativity
7
Fidelity
3
Overall
4.5
Review
The model failed completely on CSS syntax: it omitted all colons (:) after properties (e.g., 'margin 0' instead of 'margin: 0;'), rendering the code entirely invalid and non-functional in a browser. Although the aesthetic intent is good, adherence to technical constraints is non-existent because the code is syntactically incorrect, resulting in a very low fidelity score.
openai gpt-4o-mini
8.6/10 9.7 s
openai gpt-4o-mini
Tokens 518
Source code 603 B
Time 9.7 s
Matania Judgment
Animation
8
Rendering
7
Code quality
9
Creativity
5
Fidelity
10
Overall
8.63
Review
The model strictly adheres to all technical constraints, notably the absence of SVG/JS and the strict output format. The animation is fluid, combining rotation and pulsation, although the aesthetic remains very conventional (standard spinner style). The centering and code structure are perfect.
openai gpt-5.4
9.5/10 10.1 s
openai gpt-5.4
Tokens 889
Source code 2.0 KB
Time 10.1 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the absence of SVG and JavaScript, as well as the strict output format. The aesthetics are highly successful, featuring gradients and glows that provide a modern, premium look. The animations are fluid and intelligently layered to create a sense of depth.
openai gpt-5.4-mini
9.5/10 4.4 s
openai gpt-5.4-mini
Tokens 668
Source code 1.2 KB
Time 4.4 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the absence of SVG and the exclusive use of HTML/CSS. The aesthetics are very modern, featuring excellent use of gradients and shadows, and the animation combining rotation and pulsation is both fluid and elegant.
openai gpt-5.4-nano
9.6/10 11.9 s
openai gpt-5.4-nano
Tokens 1 418
Source code 4.1 KB
Time 11.9 s
Matania Judgment
Animation
9
Rendering
10
Code quality
9
Creativity
9
Fidelity
10
Overall
9.63
Review
The model perfectly adhered to all technical constraints, specifically the prohibition of SVG and JavaScript. The visual rendering is exceptional, featuring sophisticated handling of shadows, gradients, and overlays (glassmorphism), which creates a very modern sense of depth. The animation is smooth and multi-layered, offering a rich and elegant visual experience.
openai gpt-5.4-pro
9.8/10 256.6 s
openai gpt-5.4-pro
Tokens 1 262
Source code 3.5 KB
Time 256.6 s
Matania Judgment
Animation
9
Rendering
10
Code quality
10
Creativity
9
Fidelity
10
Overall
9.75
Review
The model perfectly adhered to all technical constraints, including the absence of SVG/JS and the strict output format. The rendering is exceptional, featuring a modern 'glassmorphism' aesthetic and complex, layered animations that create a very elegant visual depth. The code quality is exemplary, utilizing optimized keyframes to achieve a smooth breathing and rotation effect.
openai gpt-5.5
9.5/10 8.3 s
openai gpt-5.5
Tokens 785
Source code 1.6 KB
Time 8.3 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model perfectly adhered to all technical constraints, specifically the absence of SVG/JS and the strict output format. The aesthetics are modern, featuring a very successful use of drop shadows and harmonious colors. The combined animation (rotation and pulsation) is fluid and elegant.
openai gpt-5.5-pro
9.5/10 227.8 s
openai gpt-5.5-pro
Tokens 806
Source code 1.7 KB
Time 227.8 s
Matania Judgment
Animation
9
Rendering
9
Code quality
10
Creativity
8
Fidelity
10
Overall
9.5
Review
The model scrupulously adheres to all technical constraints, particularly the absence of SVG/JS and the strict output format. The aesthetics are modern and highly polished, featuring an intelligent use of pseudo-elements to create depth without the use of images. The animation is fluid, and the nested animations (rotation and pulsation) create a high-end visual effect.
productivia matania-latest
8.8/10 5.1 s
productivia matania-latest
Tokens 808
Source code 1.7 KB
Time 5.1 s
Matania Judgment
Animation
9
Rendering
5
Code quality
9
Creativity
7
Fidelity
10
Overall
8.75
Review
The code is excellent and strictly adheres to all technical constraints, including the absence of SVGs and the specified output format. The animation is well-conceived, featuring different speeds and directions for the rings. However, in the absence of a screenshot, the rendering score is capped by default.
xai grok-4-1-fast-non-reasoning
9.1/10 3.3 s
xai grok-4-1-fast-non-reasoning
Tokens 632
Source code 1.0 KB
Time 3.3 s
Matania Judgment
Animation
8
Rendering
9
Code quality
10
Creativity
6
Fidelity
10
Overall
9.13
Review
The model strictly adheres to all technical constraints, notably the absence of SVG and JavaScript, as well as the strict output format. The aesthetics are modern and elegant, featuring a judicious use of shadows and colors. The animation is smooth, although the effect remains fairly classic for a loader.
xai grok-4-1-fast-reasoning
9.3/10 34.5 s
xai grok-4-1-fast-reasoning
Tokens 678
Source code 1.2 KB
Time 34.5 s
Matania Judgment
Animation
9
Rendering
8
Code quality
10
Creativity
7
Fidelity
10
Overall
9.25
Review
The model strictly adheres to all technical constraints, notably the absence of SVGs and the strict output format. The animation is fluid, elegantly combining rotation and pulsation. The rendering is clean and modern, although the design remains fairly classic for a loader.
Code