Recraft's latest image generation model at ~2048px resolution with stronger composition, refined lighting, and realistic materials for print-ready and large-scale work
Settled by community votes across 2 shared challenges, with an AI judge weighing in on each.
Recraft V4 Pro
24.3
arena score
#18 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.
Stable Diffusion 3.5 Medium
15.7
arena score
#41 of 44 in Text-to-Image
Vote tally
Where the votes landed
Recraft V4 Pro
0%
win rate
Ties
0%
Stable Diffusion 3.5 Medium
0%
win rate
Shared challenges
2
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
Recraft V4 Pro
Stable Diffusion 3.5 Medium
AI Judge Analysis
Recraft V4 Pro
- + Excellent cinematic lighting and atmosphere with a coherent moon and asteroid background.
- + High level of detail in the textures of the horse's coat and the astronaut's suit.
- − Completely failed the negative constraint to have the horse on top of the astronaut.
Stable Diffusion 3.5 Medium
- + Successfully interpreted the surreal nature of the prompt.
- + Included clear astronaut suit details and a vast space background.
- − Failed the specific spatial instruction for the horse to be 'on top' of the astronaut.
- − Significant anatomical errors, including the horse having five legs and warped legs/hooves.
- − The astronaut's legs are awkwardly clipped through the horse's body.
Verdict: Both models failed the specific instruction to reverse the positions (horse on top), providing a traditional astronaut riding a horse instead. Recraft V4 Pro is the winner because it provides a high-quality, cinematic, and anatomically correct image, whereas Stable Diffusion 3.5 Medium suffered from severe anatomical distortions and incoherent leg placement.
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.”
Recraft V4 Pro
Stable Diffusion 3.5 Medium
AI Judge Analysis
Recraft V4 Pro
- + Perfect text rendering for both the main title and all detailed event information.
- + Excellent cinematic lighting and high-quality realistic textures on the jack-o-lantern.
- + Integrated the border and banner elements seamlessly into a polished composition.
- − The background feels more like a cinematic photo than a 'dark parchment poster' or 'invitation' style.
- − The jack-o-lantern is central but sits on rocks rather than being integrated into the graphic design.
Stable Diffusion 3.5 Medium
- + Captured the vintage parchment and poster aesthetic much more effectively.
- + Excellent representation of the requested thorns, webs, and twisted tree elements in the border.
- − Significant spelling errors in the title and body text (e.g., 'Halloweeen Inviloween').
- − Missing the specific '7pm' time and failed to correctly render the requested date.
- − The design feels more like a digital illustration than a 'polished, cinematic' piece.
Verdict: Recraft V4 Pro is the clear winner due to its flawless execution of the text and high-end visual realism. While Stable Diffusion 3.5 Medium followed the 'parchment' and 'poster' layout more closely, it failed significantly on text legibility and contained numerous spelling errors and missing details.
Next steps
Explore each model
Stability AI's 2.5-billion parameter Multimodal Diffusion Transformer with improvements (MMDiT-X) text-to-image model optimized for consumer hardware, featuring improved image quality, typography, and complex prompt understanding