FLUX.2 [pro] vs Reve Image 1.0

Head-to-head across 9 challenges

FLUX.2 [pro]

60.0%

win rate

Ties

15.0%

Reve Image 1.0

25.0%

win rate

60.0% 15.0% ties 25.0%

Challenge Results

Geometric Composition

Text-to-Image

“A glass cube on a wooden table. Inside the cube is a small blue sphere. On top of the cube sits a red book. A green plant is behind the cube, partially visible through the glass. Soft window light from the left.”

FLUX.2 [pro]
Reve Image 1.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent photographic realism and lighting
  • + Highly accurate refraction and reflections in the glass cube
  • + Clear adherence to the 'partially visible through glass' requirement for the plant
  • The plant is slightly more blurred than it could be, though depth of field is realistic

Reve Image 1.0

  • + Strong adherence to the spatial arrangement of the prompt
  • + Vibrant colors
  • The blue sphere is floating unnaturally inside the cube
  • The glass cube has unrealistic thick teal edges and inconsistent reflections
  • The book is disproportionately small compared to the sphere and cube

Verdict: FLUX.2 [pro] is the superior model, producing a highly realistic photograph with convincing physics, particularly in how the glass refracts the wooden table and the plant behind it. Reve Image 1.0 fails on physical coherence, showing a sphere floating in mid-air and possessing a much more 'CG' aesthetic with less accurate lighting.

Fantasy Warrior

Text-to-Image

“Close portrait of a battle-worn paladin in ornate engraved plate armor, hair braided with small beads, faint scars and dirt on the skin, warm torchlight reflecting off metal, shallow depth of field, bokeh sparks, lifelike eyes, highly detailed texture on leather straps and cloth underlayer.”

FLUX.2 [pro]
Reve Image 1.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent texture on the leather straps and burlap-style cloth underlayer.
  • + The engraving on the golden plate armor is clear and highly detailed.
  • + Strong, lifelike eye contact with realistic skin imperfections and scars.

Reve Image 1.0

  • + Intricate engraving across the entire armor set is very impressive.
  • + Effective use of warm lighting and bokeh in the background.
  • + Accurately represents the braided hair with small silver beads.
  • The character looks very young, which slightly clashes with the 'battle-worn' description.
  • The skin textures and dirt are less defined compared to the armor.
  • The eyes feel slightly glassy and less lifelike than the competitor.

Verdict: FLUX.2 [pro] is the winner as it perfectly captures the 'battle-worn' aesthetic with realistic skin textures, deep scars, and tactile leather straps. While Reve Image 1.0 features incredibly detailed armor engravings, the overall composition and character performance in FLUX.2 [pro] feel more mature and aligned with the prompt's gritty requirements.

Bald man challenge

Editing
Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After
FLUX.2 [pro]
Before After
Reve Image 1.0
0% wins 50% ties 50% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Natural hair texture and messy, realistic flow
  • + Excellent blending with the existing sideburns and beard
  • + Perfect preservation of facial features and lighting
  • The hairline on the forehead looks slightly high and thin in one spot

Reve Image 1.0

  • + Successfully added a very thick head of hair
  • + Maintains the original facial features and environment accurately
  • The hair looks like a wig resting on top of the head rather than growing from it
  • The lighting on the hair is slightly flatter than the rest of the scene
  • Coherence between the new hair and the original beard/sideburns is poor

Verdict: FLUX.2 [pro] is the winner because it integrates the new hair naturally into the existing grooming of the subject, making it look like a cohesive, real hairstyle. Reve Image 1.0 provides a very thick head of hair as requested, but the result looks artificial and lacks the realistic scalp integration and lighting seen in the other model.

Night Sky Transformation

Editing
Edit instruction

“Change the scene to night: a deep, dark sky with subtle, glistening stars visible behind the mountain.”

Before After
FLUX.2 [pro]
Before After
Reve Image 1.0
0% wins 33% ties 67% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent source preservation, keeping the town and mountains nearly identical to the original.
  • + Clear, well-defined stars as requested.
  • + Realistic night lighting that respects the silhouettes of the terrain.
  • The sky lacks a bit of depth/gradient, appearing very flatly black.

Reve Image 1.0

  • + Successfully transitions the entire scene to a 'deep, dark' atmosphere.
  • + Good preservation of the original city layout and mountain geometry.
  • + Subtle star placement matches the prompt's request for 'subtle'.
  • The overall image is slightly too dark, losing some of the fine detail in the mountain face compared to the source.
  • The stars are very sparse and almost difficult to see.

Verdict: Both models handled the editing task exceptionally well, maintaining the core architectural and geological features of the source image perfectly. FLUX.2 [pro] is the winner because it successfully brightened the stars while darkening the sky, creating more visual interest and realism, whereas Reve Image 1.0 made the entire scene a bit too dark, obscuring the majestic detail of the Matterhorn.

Over-the-top cartoon caricature

Editing
Edit instruction

“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”

Source
FLUX.2 [pro]
Reve Image 1.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Successfully incorporates many dogs into a cohesive news studio setting.
  • + The character features effectively capture the essence of the source image in a cartoon style.
  • + Strong composition with many subtle details like hockey sticks and pucks on the desk.
  • The style is more of a generic cartoon/clipart than a classic caricature.
  • The facial features are simplified significantly, losing some of the unique identity of the subject.

Reve Image 1.0

  • + Excellent preservation of the subject's facial features while applying a caricature proportion.
  • + Humorous and creative interpretation with the dog wearing a tie and headset.
  • + Strong integration of all three themes (news, hockey jersey/stick, dog) in a high-quality render.
  • The lighting is a bit overly saturated and digital.
  • The hand holding the microphone has some slight anatomical awkwardness.

Verdict: Reve Image 1.0 is the winner as it does a much better job of preserving the subject's actual likeness from the source image while transforming it into a caricature. While FLUX.2 [pro] created a charming cartoon scene, Reve Image 1.0 captured the specific humor requested by dressing the dog in professional attire and putting the anchor in a hockey jersey.

Studio Ghibli Anime Style

Editing
Edit instruction

“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”

Source
FLUX.2 [pro]
Reve Image 1.0
67% wins 0% ties 33% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent preservation of the original character's poses and expressions.
  • + Perfectly captures the requested 'Ghibli' line art and watercolor texture.
  • + Maintains high resolution and clear details while applying the stylistic filter.
  • The woman in the foreground looks slightly more generic than the source.
  • Colors are a bit washed out compared to some Ghibli films.

Reve Image 1.0

  • + Strong hand-painted feel in the background scenery.
  • + Preserves the depth of field (blur) from the original photo fairly well.
  • The faces are poorly rendered and lack the charm of the requested style.
  • Significant distortion on the man's hand and the woman's facial features.
  • The expression of the woman in red is lost to blur.

Verdict: FLUX.2 [pro] is the clear winner as it successfully translates the iconic 'distracted boyfriend' meme into a recognizable Studio Ghibli art style while maintaining the composition and expressions of the source image. Reve Image 1.0 attempts a painterly look but fails on character anatomy and facial details, resulting in a much lower-quality illustration.

Neutral Expression to Genuine Smile

Editing
Edit instruction
{
  "action": "image_edit",
  "reference": "uploaded neutral portrait",
  "change": "Warm genuine Duchenne smile: lips curved up, slight natural teeth, soft eye crinkles, subtle cheek raise",
  "details": "Realistic smiling skin (dimples if present, soft cheek shadows), slightly brighter eyes; keep exact eye shape/color/iris",
  "preserve_exact": "Face identity/structure, eyes/nose/lips/eyebrows, hair, skin texture/pores/freckles, makeup, clothing, head pose, background, lighting, shadows, framing",
  "no_changes": "No face shape change, no new features, no gaze shift, no hair/clothing/lighting/background edits",
  "style": "Ultra-photorealistic 8K portrait, sharp face focus, natural soft lighting, realistic skin glow"
}
Before After
FLUX.2 [pro]
Before After
Reve Image 1.0
80% wins 20% ties 0% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent Duchenne smile with realistic eye crinkles and cheek elevation
  • + Significant but natural change to the facial expression
  • + Preserves hair and background almost perfectly
  • Slightly softens the skin texture compared to the original
  • Minor change to the jawline/chin shape to accommodate the larger smile

Reve Image 1.0

  • + Exceptional preservation of skin texture and freckles from the original
  • + Subtle and realistic lip/teeth modification
  • + Maintains exact identity and facial structure
  • The smile is a bit stiff and lacks the 'Duchenne' eye crinkles requested
  • The eyes remain quite neutral despite the 'brighter eyes' and 'soft eye crinkles' instruction

Verdict: Both models performed excellent edits, but FLUX.2 [pro] followed the specific 'Duchenne smile' instruction much better by involving the eyes and cheeks in the expression. Reve Image 1.0 was superior at preserving every single pore and freckle from the source, but the resulting expression feels less genuine and misses the requested eye crinkles.

Golden Hour Stroll

Editing
Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

Source
FLUX.2 [pro]
Reve Image 1.0
50% wins 0% ties 50% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent preservation of the original face and dog details.
  • + Adds leaves while maintaining a very clean background and foreground.
  • + The hair motion is dramatic and clearly follows the prompt.
  • The flying leaves look a bit static and lack motion blur.
  • A leaf is awkwardly overlapping the dog's mouth.

Reve Image 1.0

  • + Stronger sense of 'dynamic motion' through the use of motion blur on the leaves.
  • + Great hair flow that feels integrated with the wind direction.
  • + Excellent preservation of the subject's identity and the surrounding environment.
  • A few leaves appear slightly low-resolution in the blurred foreground.
  • The density of leaves might be a bit overwhelming for some tastes.

Verdict: Both models did an exceptional job of preserving the source image while applying the requested edits. FLUX.2 [pro] created a very clean image with distinct flying leaves, but Reve Image 1.0 felt more 'dynamic' and 'lively' because it included motion blur on the leaves, which better captured the requested energetic feel. FLUX.2 [pro] also had a minor issue with a leaf clipping into the dog's mouth.

Apollo 11: Journey to Tranquility

Text-to-Image

“Create a clean, modern vector infographic poster about the Apollo 11 mission. NASA-inspired palette (navy, white, muted red, light gray). Flat-vector style, crisp lines, consistent iconography, subtle gradients only. Steps (stop at landing): 1. Launch (Saturn Vicon) 2. Earth Orbit (Earth + orbit ring icon) 3. Translunar (trajectory arc icon) 4. Lunar Orbit (Moon + orbit ring icon) 5. Descent (lunar module descending icon) 6. Landing (lunar module on the surface icon) Small supporting elements (minimal text): • Crew strip: three silhouette icons with only last names: Armstrong, Aldrin, Collins. • Landing site marker: Moon pin labeled "Tranquility" only. Layout constraints: generous margins, large readable labels, clean background with subtle stars. Vector-only, print-poster look, high resolution.”

FLUX.2 [pro]
Reve Image 1.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [pro]

  • + Excellent typography and readability for main titles.
  • + Cohesive narrative layout using visual connectors between steps.
  • + Stronger vector art style with professional shading and composition.
  • Missed the numerical step numbering requested.
  • Body text under titles is garbled gibberish.

Reve Image 1.0

  • + Adhered better to the specific step-by-step numbering requested.
  • + Closer adherence to the simplified flat-vector requirement.
  • + Included the supporting text 'Tranquility' at the correct final stage.
  • Several spelling errors in header and step titles (e.g., 'Dessent', 'Lunnr').
  • The composition feels slightly cluttered and less 'premium' than Model A.
  • Inconsistent number of steps displayed compared to the prompt's list.

Verdict: FLUX.2 [pro] produced a much more professional and aesthetically pleasing infographic that looks like a finished piece of graphic design, despite missing the specific step numbers. Reve Image 1.0 followed the structural instructions better but suffered from poor spelling and a less sophisticated visual style. FLUX.2 [pro] is the winner for its superior layout, cleaner lines, and significantly better text rendering on the primary headers.

FLUX.2 [pro]

Black Forest Labs' state-of-the-art image generation model with maximum quality and speed, supporting text-to-image and multi-reference image editing with up to 4MP output

Reve Image 1.0

Reve AI's text-to-image generation model with strong aesthetic quality, accurate text rendering, and detailed instruction following capabilities