Bria's 8B parameter model for controlled text-to-image generation with structured JSON prompts, trained exclusively on licensed data
Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.
Bria FIBO
#55 of 62 in Text-to-Image
Qwen Image Edit 2509
#32 of 32 in Image Editing
Where the votes landed
Bria FIBO
0%
win rate
Ties
0%
Qwen Image Edit 2509
0%
win rate
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Bald man challenge
Image Editing“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
AI Judge Analysis
Bria FIBO
- + The hair rendering is highly realistic with high-quality textures.
- + The head shape and neck area were adjusted to accommodate the new hair style.
- − Completely failed to preserve the source image features, changing the face, clothing details, and background environment.
- − Does not look like the original person at all.
Qwen Image Edit 2509
- + Excellent source preservation, maintaining identical facial features, clothing, and background.
- + Applied the edit exactly as requested while keeping the original context intact.
- − The hair texture has a slightly 'wavy' or digitally generated sheen that looks less natural than others.
- − The blending at the temples where the hair meets the existing beard is a bit abrupt.
Verdict: Qwen Image Edit 2509 is the clear winner as it successfully performed the edit while perfectly preserving the identity of the person and the background, whereas Bria FIBO generated an entirely new image that barely resembles the source. Despite Bria FIBO having slightly more realistic-looking hair fibers in isolation, it failed the fundamental requirements of an image editing task.
Over-the-top cartoon caricature
Editing“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.”
AI Judge Analysis
Bria FIBO
- + Excellent caricature style with highly exaggerated facial features
- + Successfully incorporates a dog and a news anchor desk setting
- − Completely fails to include the hockey element requested
- − Source preservation is weak as the subject's face is barely recognizable
Qwen Image Edit 2509
- + Successfully incorporates all elements: news anchor desk, dog, and hockey players/rink
- + Preserves the subject's features and clothing (denim shirt) much better while applying a caricature effect
- + Maintains the composition and 'selfie' angle of the source image
- − The text in the speech bubbles is gibberish
- − The dog is very small and lacks detail compared to the rest of the image
Verdict: Qwen Image Edit 2509 is the clear winner because it followed all parts of the multi-prompt instruction, including the hockey element which Bria FIBO completely ignored. Qwen also did a much better job of preserving the identity and wardrobe of the person in the source image while still achieving a fun caricature style.
Studio Ghibli Anime Style
Editing“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”
AI Judge Analysis
Bria FIBO
- + Excellent application of the 'dreamy' Ghibli aesthetic with soft lighting and flowers.
- + High artistic quality for the background and textures.
- + Accurately captured the requested pastel color palette.
- − Failed to preserve the composition as the main male character is now walking towards the camera rather than looking back.
- − Significant change to the character poses and background architecture.
Qwen Image Edit 2509
- + Excellent source preservation, maintaining the exact poses and composition of the original meme.
- + Successfully translated the facial expressions into an anime/illustration style.
- + Preserved the specific background elements like the red bus and street layout.
- − The art style leans more towards generic manhwa or modern digital illustration than specific Studio Ghibli painterly textures.
- − The color palette is a bit washed out rather than 'warm and nostalgic'.
Verdict: Qwen Image Edit 2509 is the clear winner for an editing task because it perfectly preserved the identity and composition of the famous 'distracted boyfriend' meme while applying the requested style change. Bria FIBO created a beautiful illustration that captured the Ghibli mood better, but it failed as an edit by completely altering the characters' poses and the overall scene structure.
Golden Hour Stroll
Image Editing“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
AI Judge Analysis
Bria FIBO
- + Successfully captures a sense of dynamic movement and energy throughout the image.
- + Significantly changes the pose and hair to reflect the requested windy atmosphere.
- + Integrates leaves well into the overall scene's lighting and motion.
- − Fails as an edit by regenerating the entire image, losing the original's likeness and background details.
- − The dog's breed and features changed notably from the source.
Qwen Image Edit 2509
- + Excellent source preservation, keeping the original woman, dog, and background almost perfectly intact.
- + Successfully adds the requested blowing hair effect while maintaining facial likeness.
- + Leaves are scattered across the scene as requested.
- − The added leaves look like stickers overlaid on the image rather than part of a 3D space.
- − The hair rendering is a bit computerized and lacks natural texture compared to the rest of the image.
Verdict: Bria FIBO completely regenerated the image, failing the primary goal of image editing which is to preserve the source content. Qwen Image Edit 2509 successfully applied the 'wind' and 'leaves' instructions while keeping the original subject and environment perfectly intact, making it the effective winner for this task.
Explore each model
Alibaba's Qwen image editing model for instruction-based image modifications and transformations