ARENA Leaderboard
See how AI image models stack up against each other. How it works
Which model makes the best edits?
Same source image, same instruction, blind community votes. See which models handle edits best.
Best AI Models for Image Upscaling
| # | Model | Elo |
|---|---|---|
| 1 | 1255 | |
| 2 | 1150 | |
| 3 | 1125 | |
| 4 | 1036 |
Google
As of June 2026, Crystal Upscaler dominates the leaderboard with a 1255 Elo and a 95.9% win rate, maintaining a significant 105-point gap over the second-place Recraft Crisp Upscale (1150 Elo). While the top spot is uncontested, competition for the runner-up position remains tight, with only 25 Elo points separating Recraft from Topaz Image Upscale (1125 Elo). Notably, Recraft Crisp Upscale secures the second rank as a high-value budget option, offering speeds 70% faster and a price point 75% lower per image than the leader.
Elo vs Cost
Elo vs Speed
Speed data is still warming up
We only have enough recent requests for Crystal Upscaler (30.2s average).
Challenges
Feather Fidelity Challenge Image Upscaling
Crystal Upscaler
Recraft Crisp Upscale
Freckle Fortress Challenge Image Upscaling
Crystal Upscaler
Recraft Crisp Upscale
Tokaji Text Trials Image Upscaling
Crystal Upscaler
Recraft Crisp Upscale
Weave Warfare Image Upscaling
Crystal Upscaler
Topaz Image Upscale
Recraft Crisp Upscale
FAQ
What is the best AI image upscaling model?
Based on blind community voting, Crystal Upscaler is currently the #1 ranked AI image upscaling model with an Elo rating of 1255. Rankings update in real time as new votes come in.
How are AI image upscaling models ranked on Lumenfall?
Lumenfall Arena ranks AI models through blind community voting. In each matchup, two models generate from the same prompt and voters pick the better result without seeing model names. Votes are processed using TrueSkill, a Bayesian rating algorithm developed by Microsoft Research, that produces a single Elo score reflecting each model's relative quality.
What is an Elo rating for AI models?
An Elo rating is a numerical score representing a model's skill relative to other models. Under the hood, Lumenfall uses TrueSkill, which tracks two values per model: mu (estimated skill) and sigma (uncertainty). The displayed Elo is calculated as 1000 + 10 x (mu - 3*sigma), a conservative lower bound. A model must prove itself consistently across many matchups to earn a high rating.