Generation highlights
How the top models compare
Best models, by
Best Image Editing Models by Price
1 model without pricing omitted.
Best Image Editing Models by Speed
14 models waiting for enough speed data.
Best AI Models for Image Editing
| # | Model | Elo |
|---|---|---|
| 1 | 1250 | |
| 2 | 1239 | |
| 3 | 1235 | |
| 4 | 1230 | |
| 5 | 1223 | |
| 6 | 1221 | |
| 7 | 1221 | |
| 8 | 1216 | |
| 9 | 1214 | |
| 10 | 1214 |
| 11 | 1212 | |
| 12 | 1209 | |
| 13 | 1208 | |
| 14 | 1207 | |
| 15 | 1205 | |
| 16 | 1205 | |
| 17 | 1199 | |
| 18 | 1195 | |
| 19 | 1190 | |
| 20 | 1189 | |
| 21 | 1188 | |
| 22 | 1186 | |
| 23 | 1185 | |
| 24 | 1180 | |
| 25 | 1173 | |
| 26 | 1170 | |
| 27 | 1167 | |
| 28 | 1166 | |
| 29 | 1158 | |
| 30 | 1146 |
As of October 2026, Google’s Nano Banana Pro leads the Image Editing leaderboard with a 1250 Elo, maintaining a narrow 11-point lead over its sibling model, Nano Banana 2. Competition at the top remains fierce, as GPT Image 2 follows closely at 1235 Elo with a dominant 84.2% win rate despite its high latency. Notably, Nano Banana 2 Lite secures a top-five position at 1223 Elo, offering a 69.6% win rate and fast generation speeds for 75% less cost than the top-ranked model.
Rivalries
Aggregate head-to-head across the arena
Highlighted challenges
Man and Car in California
Combines two references (a person and a car) into one believable driving scene. It checks whether the model keeps the man’s identity and the car’s appearance while placing them convincingly on the coastline with matched lighting and perspective.
Bald man challenge
A targeted edit test. Adding a realistic head of hair requires a believable hairline and texture while keeping the face, expression, and lighting unchanged. Common failures are wig-like or pasted-on hair and subtle changes to the underlying face.
Recent SOTA shifts in Image Editing
Full historyFAQ
What is the best AI image editing model?
Based on blind community voting, Nano Banana Pro is currently the #1 ranked AI image editing model with an Elo rating of 1250. Rankings update in real time as new votes come in.
How are AI image editing 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.
Keep the arena honest
Cast your vote
Pick winners in blind matchups. Every vote nudges the Elo and shapes these rankings.
Cast Your VoteSuggest a prompt
Got an idea worth testing? Submit a prompt and watch the models battle it out.