Square 1:1 photograph of two hands passing a full espresso cup on a saucer from one to the other, four fingers of each hand visible around the saucer's rim, thumbs on top. Steam rises from the cup, nothing spills. Warm café light, marble table.

Hands & Gestures · default settings, one run, 1:1 · Updated September 20, 2026

Not run yet
Z Image Pro Uncensored
Nano Banana 2.1 Advanced: Passing the Espresso
Nano Banana 2.1 Advanced
4.4imagemodels.ai score
  • Prompt adherence4.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The two-handed espresso handoff has convincing steam and café warmth.

Not run yet
Nano Banana 2.1
Not run yet
Qwen Image 2.1
Not run yet
Qwen Image 3.0
Alibaba Qwen Image 3.0 Pro: Passing the Espresso
Alibaba Qwen Image 3.0 Pro
3.9imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.0
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Warm, polished handoff image, but it introduces multiple unrequested text elements.

Qwen Image Max Uncensored: Passing the Espresso
Qwen Image Max Uncensored
3.4imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Three hands and a landscape crop miss the specified two-hand square handoff.

Nano Banana Pro: Passing the Espresso
Nano Banana Pro
3.6imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Convincing steamy espresso photo, but the exact finger placement is not met.

Nano Banana 2: Passing the Espresso
Nano Banana 2
4.1imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

A strong square saucer handoff, with only the exact four-finger arrangement falling short.

GPT Image 2: Passing the Espresso
GPT Image 2
4.2imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.5
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

A polished, stable handoff with attractive café light, but the requested finger visibility is incomplete.

GPT Image 2.5 Flare: Passing the Espresso
GPT Image 2.5 Flare
4.0imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Richly lit and polished, but the cup reads too large and the mandated grip is not shown.

GPT Image 2.5 Sunburst: Passing the Espresso
GPT Image 2.5 Sunburst
4.1imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Beautifully composed and physically stable, though finger visibility does not meet the exact requirement.

Seedream 5.0 Pro: Passing the Espresso
Seedream 5.0 Pro
3.9imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The café handoff is polished, but the required four visible fingers per hand are absent.

Grok Imagine Image V2: Passing the Espresso
Grok Imagine Image V2
4.1imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity5.0
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Strong photoreal café handoff, but the exact four-fingers-per-hand saucer grip is not visible.

FLUX.2 Pro: Passing the Espresso
FLUX.2 Pro
4.1imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The exchange reads well, though the exact four-fingers-per-hand arrangement is obscured.

P-Image Ideogram: Passing the Espresso
P-Image Ideogram
3.9imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The handoff is strong, but the specified four visible fingers per hand are not clear.

Qwen Image: Passing the Espresso
Qwen Image
3.9imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.0
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The transfer reads well, but the explicit finger count and thumbs-on-top pose are missed.

Z-Image Turbo: Passing the Espresso
Z-Image Turbo
3.6imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.0
  • Quality & detail4.0
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.0
  • Text rendering—

It captures the scene, but neither hand shows the requested four fingers around the rim.

HiDream I1 Full: Passing the Espresso
HiDream I1 Full
3.4imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Polished cafe photo, but both hands grip the cup rather than pass the saucer.

Nano Banana: Passing the Espresso
Nano Banana
3.7imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Strong steam and cafe setting, but the hands do not show the specified saucer-rim grip.

Nano Banana 2 Lite: Passing the Espresso
Nano Banana 2 Lite
3.9imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The saucer handoff reads clearly, though the required finger count and pose are not exact.

GPT Image 1.5: Passing the Espresso
GPT Image 1.5
4.1imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Strong warm photo and convincing steam, with the precise saucer-rim finger count still missed.

Seedream 5.0 Lite: Passing the Espresso
Seedream 5.0 Lite
3.3imagemodels.ai score
  • Prompt adherence2.0
  • Style fidelity4.0
  • Quality & detail4.0
  • Realism & anatomy4.0
  • Composition & framing3.0
  • Look & lighting4.0
  • Text rendering—

The wide crop and incorrect finger placement miss two explicit requirements.

Seedream 4.5: Passing the Espresso
Seedream 4.5
3.8imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Strong scene rendering, but it is landscape and lacks the required visible finger arrangement.

Grok Imagine Image: Passing the Espresso
Grok Imagine Image
4.1imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

A strong café handoff image, held back by crowded, indistinct fingers around the saucer.

FLUX.2 Flex: Passing the Espresso
FLUX.2 Flex
4.6imagemodels.ai score
  • Prompt adherence4.0
  • Style fidelity5.0
  • Quality & detail5.0
  • Realism & anatomy4.5
  • Composition & framing4.5
  • Look & lighting5.0
  • Text rendering—

Excellent polished square handoff with steam and marble, though the required full finger count is still obscured.

FLUX 1.1 Pro Ultra: Passing the Espresso
FLUX 1.1 Pro Ultra
4.1imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity5.0
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Clean square café composition, but both hands use cup-side grips instead of the specified saucer-rim hold.

P-Image: Passing the Espresso
P-Image
3.2imagemodels.ai score
  • Prompt adherence1.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.0
  • Composition & framing4.0
  • Look & lighting4.0
  • Text rendering—

Four hands and two cups replace the requested two-hand single-cup transfer.

Microsoft Lens: Passing the Espresso
Microsoft Lens
2.9imagemodels.ai score
  • Prompt adherence1.0
  • Style fidelity4.0
  • Quality & detail4.0
  • Realism & anatomy2.5
  • Composition & framing3.5
  • Look & lighting4.5
  • Text rendering—

It has an attractive cafe look, but three hands and a latte violate the requested setup.

SenseNova U1 8B: Passing the Espresso
SenseNova U1 8B
3.0imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.0
  • Quality & detail3.5
  • Realism & anatomy3.0
  • Composition & framing3.5
  • Look & lighting4.0
  • Text rendering—

The non-square image uses an oversized coffee cup and obscured incorrect hand placement.

ERNIE Image: Passing the Espresso
ERNIE Image
3.5imagemodels.ai score
  • Prompt adherence2.0
  • Style fidelity4.5
  • Quality & detail4.0
  • Realism & anatomy4.0
  • Composition & framing3.5
  • Look & lighting4.5
  • Text rendering—

Warm realistic coffee scene, but it is portrait on wood rather than square on marble.

ERNIE Image Turbo: Passing the Espresso
ERNIE Image Turbo
3.6imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.0
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Attractive café image, but portrait framing and the implausible two-handled grip miss the exact pass.

GLM Image: Passing the Espresso
GLM Image
4.2imagemodels.ai score
  • Prompt adherence3.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy4.5
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Convincing handoff, materials, and warm café mood; only the exact finger-count instruction is missed.

FLUX.2 Klein 9B: Passing the Espresso
FLUX.2 Klein 9B
3.7imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

Excellent café photo, but it holds the cup rather than showing the required saucer-rim handoff.

FLUX.2 Klein 4B: Passing the Espresso
FLUX.2 Klein 4B
3.7imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Atmospheric and clean, but the saucer remains on the table and the requested grip is absent.

FLUX.2 Dev: Passing the Espresso
FLUX.2 Dev
4.4imagemodels.ai score
  • Prompt adherence4.0
  • Style fidelity5.0
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting4.5
  • Text rendering—

The strongest overall match, with a convincing square marble-table handoff; exact finger count remains obscured.

FLUX.2 Dev Turbo: Passing the Espresso
FLUX.2 Dev Turbo
3.4imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity5.0
  • Quality & detail4.0
  • Realism & anatomy4.0
  • Composition & framing3.5
  • Look & lighting4.0
  • Text rendering—

The cup, steam, and marble are convincing, but four hands plainly violate the requested two-hand pass.

LongCat Image: Passing the Espresso
LongCat Image
3.6imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.0
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing3.0
  • Look & lighting4.5
  • Text rendering—

The handoff and steam work well, but the portrait framing misses the required square format.

Wan Image Small: Passing the Espresso
Wan Image Small
3.2imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.0
  • Quality & detail4.0
  • Realism & anatomy3.5
  • Composition & framing3.5
  • Look & lighting4.0
  • Text rendering—

An extra hand and saucer make the transfer count and arrangement incorrect.

Hunyuan FLUX SRPO: Passing the Espresso
Hunyuan FLUX SRPO
3.6imagemodels.ai score
  • Prompt adherence2.0
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The image is clean and atmospheric, but the left hand holds the cup instead of the saucer.

Seedream 4.0: Passing the Espresso
Seedream 4.0
3.3imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.5
  • Quality & detail4.0
  • Realism & anatomy3.5
  • Composition & framing3.5
  • Look & lighting4.5
  • Text rendering—

Landscape framing and a third hand break the exact requested setup.

HiDream I1 Fast: Passing the Espresso
HiDream I1 Fast
3.4imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The warm photo is well rendered, but it shows a large latte held by the cup.

HiDream I1 Dev: Passing the Espresso
HiDream I1 Dev
3.4imagemodels.ai score
  • Prompt adherence1.5
  • Style fidelity4.5
  • Quality & detail4.5
  • Realism & anatomy3.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

The image has convincing steam and light, but the hands grip a large latte rather than the saucer.

FLUX.1 Krea Dev: Passing the Espresso
FLUX.1 Krea Dev
4.0imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity5.0
  • Quality & detail4.5
  • Realism & anatomy4.0
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Rich, sharp café rendering, but the gesture reads as reaching or adjusting rather than a saucer-rim pass.

FLUX 1.1 Pro: Passing the Espresso
FLUX 1.1 Pro
4.0imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity5.0
  • Quality & detail4.5
  • Realism & anatomy4.5
  • Composition & framing4.0
  • Look & lighting4.5
  • Text rendering—

Beautiful realistic scene, but the landscape crop and handle grip miss the specified saucer pass.

FLUX.1 Dev: Passing the Espresso
FLUX.1 Dev
4.0imagemodels.ai score
  • Prompt adherence2.5
  • Style fidelity4.5
  • Quality & detail5.0
  • Realism & anatomy4.0
  • Composition & framing4.5
  • Look & lighting5.0
  • Text rendering—

Exceptionally polished warm image, but it reads as cup-handle holding rather than a saucer-rim handoff.

FLUX.1 Schnell: Passing the Espresso
FLUX.1 Schnell
3.6imagemodels.ai score
  • Prompt adherence3.0
  • Style fidelity4.5
  • Quality & detail4.0
  • Realism & anatomy3.0
  • Composition & framing4.0
  • Look & lighting4.0
  • Text rendering—

Warm café scene works, but the specified saucer-rim finger arrangement is not clear.