HiDream AI's image-to-image editing model for instruction-based image modifications and transformations
Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.
HiDream E1
0.7
arena score
#33 of 32 in Image Editing
Skill signature
· Image Editing
Wan 2.6
23.3
arena score
#28 of 62 in Text-to-Image
Top 2 in Image-to-Video
Vote tally
Where the votes landed
HiDream E1
0%
win rate
Ties
0%
Wan 2.6
0%
win rate
Shared challenges
4
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Bald man challenge
Image EditingEdit instruction
“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
HiDream E1
Wan 2.6
AI Judge Analysis
HiDream E1
- + Successfully added high-volume, thick hair.
- − Completely altered the background from a scrubland to a desert with dunes.
- − Distorted the person's face and jacket texture with digital artifacts.
- − Hair looks like a synthetic wig rather than natural hair.
Wan 2.6
- + Excellent source preservation, keeping the background and clothing identical to the original.
- + The hair has a very realistic texture and a natural-looking hairline.
- + Maintains the original lighting and skin details perfectly.
- − Slightly altered the shape of the ear on the left side to accommodate the sideburns.
Verdict: Wan 2.6 is the clear winner as it perfectly executed the head-swap edit while preserving every other element of the original image, including the background and lighting. HiDream E1 failed as an editing tool, regenerating the entire scene into a lower-quality image with a different environment and distorted facial features.
Over-the-top cartoon caricature
EditingEdit 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
HiDream E1
Wan 2.6
AI Judge Analysis
HiDream E1
- + Expertly blends the three themes into a single scene with a hockey rink background.
- + Maintains a higher level of facial resemblance to the source subject through the caricature distortion.
- + Features a well-executed 3D render style with vibrant lighting.
- − Significant issues with hand anatomy, including extra and fused fingers.
- − The dog's integration with the arms is physically confusing/clipping.
Wan 2.6
- + Comprehensive professional setting with a news desk and broadcast lighting.
- + Includes high-quality cartoon-style dogs with great character.
- + Excellent execution of the hockey prop and accessories like the headset and microphone.
- − The facial features in the cartoon style lose almost all resemblance to the original woman.
- − The hockey stick is held in a somewhat awkward manner.
Verdict: HiDream E1 does a better job of preserving the subject's identity within the caricature, successfully morphing her specific features into a humorous 3D style. However, Wan 2.6 provides a much more coherent and detailed 'TV news' composition and avoids the severe anatomical glitches (hand artifacts) found in the other model, making it the more polished final illustration despite the loss of facial resemblance.
Studio Ghibli Anime Style
EditingEdit 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
HiDream E1
Wan 2.6
AI Judge Analysis
HiDream E1
- + Accurately captures the character composition and poses from the source.
- + Adopts a clear 2D anime style with clean lines.
- − Fails to preserve the colors of the source image, turning the red dress and blue plaid shirt into generic pink tones.
- − Visual style is more generic modern webcomic than the requested Ghibli hand-painted aesthetic.
Wan 2.6
- + Excellent preservation of original colors, including the red dress and blue plaid pattern.
- + Perfectly captures the Studio Ghibli 'watercolor' and hand-painted texture request.
- + Maintains the specific facial expressions and lighting of the original while stylizing them.
- − Some minor white speckle artifacts across the image.
Verdict: Wan 2.6 is the clear winner as it successfully follows the complex 'Studio Ghibli' aesthetic instructions while preserving the key elements of the source image like the red dress and plaid shirt. HiDream E1 transforms the image into a generic anime style and loses too much information from the original photo, such as the specific colors of the clothing.
Golden Hour Stroll
Image EditingEdit instruction
“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
HiDream E1
Wan 2.6
AI Judge Analysis
HiDream E1
- + Excellent depiction of wind-blown hair
- + Changes the leaf colors to a dynamic autumn red for contrast
- − The leaf assets look like flat overlays rather than naturally integrated objects
- − Slightly alters the facial features and lighting compared to the source
Wan 2.6
- + Near perfect lighting and color consistency with the source image
- + Leaves are integrated more naturally into the 3D space of the scene
- − The motion in the hair is much more subtle than requested
- − Added leaves are small and feel less 'energetic' than the prompt suggested
Verdict: HiDream E1 followed the instructions for a dynamic feel much more effectively, creating a clear sense of movement in the hair and a bold visual with the flying leaves. Wan 2.6 is technically superior at preserving the source image's identity and lighting, but it was too conservative with the requested edit, resulting in a less lively image.
Next steps
Explore each model
Alibaba's multimodal generation model from the Wan AI suite, supporting text-to-video, image-to-video, reference-to-video with audio, and text-to-image, in both Chinese and English