FP8 quantized variant of Black Forest Labs' FLUX.1 [schnell] model, offering ~2x faster inference with reduced precision while maintaining high-quality image generation in 4 steps
Settled by community votes across 3 shared challenges, with an AI judge weighing in on each.
FLUX.1 [schnell] FP8
17.8
arena score
#36 of 44 in Text-to-Image
Skill signature
Not enough comparable category data
The chart appears once both models have ratings across at least three shared arena categories.
GPT Image 1.5
26.5
arena score
#7 of 44 in Text-to-Image
Top 3 in Image Editing
Vote tally
Where the votes landed
FLUX.1 [schnell] FP8
0.0%
win rate
Ties
100.0%
GPT Image 1.5
0.0%
win rate
0.0%
100.0% ties
0.0%
Shared challenges
3
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Chalkboard Menu
Text-to-Image“Handwritten-style chalkboard menu in a cozy café, all text rendered in the exact same realistic chalk handwriting style with natural variations in letter size, slight slant, and chalk texture — no printed or digital fonts anywhere on the board. Title at the top in elegant cursive chalk handwriting: ‘TODAY’S SPECIALS – APRIL 30, 2026’. Below it, three menu items also in the same handwritten chalk style: ‘Truffle Mushroom Risotto – $24’, ‘Grilled Octopus with Lemon & Herbs – $28’, ‘Brown Butter Chocolate Chip Cookies – $9’. At the very bottom, smaller text in the identical handwritten chalk style (slightly smaller but still clearly legible with the same handwriting characteristics): ‘All items made fresh daily • Ask about our gluten-free options’. Warm ambient café lighting, visible chalk dust and smudges, realistic handwriting imperfections, no clean printed text anywhere.”
FLUX.1 [schnell] FP8
GPT Image 1.5
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Crisp image quality and high resolution
- + Captures the aesthetic of a wooden-framed chalkboard effectively
- − Significant spelling errors and repeated words across the menu
- − Failure to use cursive style for the title as requested
- − Text rendering resembles a digital marker font rather than gritty chalk texture
GPT Image 1.5
- + Perfect adherence to spelling and specific menu items
- + Excellent chalk texture and realistic handwritten cursive style
- + Accurate layout following all prompt instructions including the date and prices
- − The frame of the chalkboard is slightly cropped at the edges
- − The lighting at the very top is a bit harsh
Verdict: GPT Image 1.5 is the clear winner as it followed all complex text instructions perfectly, including specific spelling, prices, and the requirement for cursive handwriting. FLUX.1 [schnell] FP8 struggled significantly with the text, producing numerous spelling errors and failing to render the title in cursive as requested.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.1 [schnell] FP8
GPT Image 1.5
0% wins
100% ties
0% wins
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Successfully followed the difficult spatial instruction of having the horse on top of the astronaut/equipment.
- + High visual clarity and striking cinematic lighting.
- + Creative interpretation of a surreal concept including multiple horse features and space-tech.
- − Anatomy is a bit messy with multiple horse heads and necks emerging from the central figure.
- − The astronaut is represented more as a piece of equipment than a person.
GPT Image 1.5
- + Excellent anatomical detail on the horse and space suit.
- + Rich, busy background with impressive celestial objects and surface textures.
- + High level of technical detail in the harness and lunar lander.
- − Failed the primary spatial prompt instruction: 'horse on top, not vice versa'.
- − Followed the common cliche instead of the specific surreal request.
Verdict: FLUX.1 [schnell] FP8 is the clear winner because it actually attempted the 'horse on top' instruction, resulting in a surreal and unique composition. GPT Image 1.5 produced a much higher quality image in terms of detail and texture, but it completely ignored the specific spatial constraint to follow a standard 'astronaut riding horse' trope.
The Halloween Invitation
Text-to-Image“Vintage gothic Halloween party invitation. Dark parchment poster, spooky border with webs and thorns, central glowing jack-o-lantern, bats, twisted trees, moody night sky. Add elegant gothic title text saying "Halloween Party Invitation", a small scroll banner saying "You are invited to a night of frights", and event details at the bottom: Date: 30.10.2026 Time: 7pm Location: The Arches, NYC Spooky but polished, cinematic lighting, square format.”
FLUX.1 [schnell] FP8
GPT Image 1.5
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Features a bold, central jack-o-lantern.
- + Captures the 'moody night sky' with a clean silhouette effect.
- − Several severe spelling errors in the banner and event details (e.g., 'A ai nigh tof friights', 'Theaches').
- − The border looks like a flat gradient rather than the requested 'webs and thorns'.
- − Text layout is cluttered and uses inconsistent fonts.
GPT Image 1.5
- + Excellent adherence to all text requirements with perfect spelling.
- + Highly detailed 'webs and thorns' border that matches the vintage gothic aesthetic.
- + Superior composition with a more natural integration of the pumpkin into the environment.
- − The layout is a bit dense with the background details.
- − The jack-o-lantern is slightly off-center.
Verdict: GPT Image 1.5 is the clear winner as it followed every complex text instruction perfectly while maintaining a high-quality vintage aesthetic. FLUX.1 [schnell] FP8 struggled significantly with the typography, resulting in numerous spelling errors and a much less cohesive gothic border.
Next steps
Explore each model
OpenAI's state-of-the-art image generation model with better instruction following and adherence to prompts