ENفا
Get API key
MODELSArena’s interaction judge tests what agents actually finish▼ 22%DeepSeek V4.1 Flash API price#1Claude Fable 5.2 · 30.48% win shareFUNDINGAnthropic targets $64.4B Nasdaq IPO as revenue surgesTOOLSCursor tops the tools directory this weekMODELSGoogle cuts Gemini 4 Carbon prices before Gemini 2.5 sunsetPOLICYFive frontier AI labs face coordinated SF safety lawsuits
Vigas Arena · updated just now

The best Image edit models,
ranked by real use.

Compares spatial prompt adherence, typography consistency, object removal, and style-preserving image edits.

378,706,203sessions rankedHow we rank →

Rising this week · Image edit

GPT Image 2.5 Sunburstup 3.27 pts · #1 in Image editGemini Nano Banana 2.1up 1.55 pts · #3 in Image editGemini 4 Carbonup 1.53 pts · #5 in Image edit

The podium · Image edit

Win share · this week
  1. Rank 1: GPT Image 2.5 SunburstOpenAI · 1519 score24.80%
  2. Rank 2: GPT Image 2.5 FlareOpenAI · 1484 score22.30%
  3. Rank 3: Gemini Nano Banana 2.1Google · 1427 score20.40%

Full ranking

Image edit leaderboard: win share, score, cost per task and votes
#ModelVotes7 days
1GPT Image 2.5 SunburstOpenAI24.80%1519 ±5$1.80112,000
2GPT Image 2.5 FlareOpenAI22.30%1484 ±4$1.1098,733
3Gemini Nano Banana 2.1Google20.40%1427 ±6$0.4591,466
4Muse SparkMeta18.10%1408 ±7$0.6284,199
5Gemini 4 Carbon (High)Google16.50%1395 ±8$1.4279,932
6Qwen 3.8 MaxAlibaba14.90%1382 ±9$0.5472,665
Score 1519 ±5$1.80/task112,000 votes
#222.30%
Score 1484 ±4$1.10/task98,733 votes
Score 1427 ±6$0.45/task91,466 votes
#418.10%
Score 1408 ±7$0.62/task84,199 votes
Score 1395 ±8$1.42/task79,932 votes
#614.90%
Score 1382 ±9$0.54/task72,665 votes

Pareto frontier

score vs $/task

Green models give the best score for their price. Tap any dot to inspect value vs. performance.

GPT Image 2.5 SunburstPareto optimal
Score 1519 · 24.80% win · $1.80/task
API pricing

Live sessions

LIVE

No sessions in the last 24 hours yet. Send a request through the Vigas API.

Connect via API — start in 30s
§ Methodology

How the rankings are made.

Real sessions, not benchmarks

Win share counts how often a model finishes real tasks that users accept — across real API workloads, coding agents, and blind Arena tests.

Blind pairwise votes

In Arena, people compare two hidden models and vote. Scores use a Bradley–Terry model with 95% confidence intervals.

Cost is part of the score

Every ranking shows dollars per completed task, so you can see who is best for the money — not just best.

§ Image edit · Questions & explore

About the image edit board.

AI image generatorsTools newsAPI pricesAPI quickstart

What is the #1 AI model on the Image edit leaderboard right now?

GPT Image 2.5 Sunburst (OpenAI) leads image edit with a 24.80% real-work win share and a Bradley-Terry score of 1519, ahead of #2 GPT Image 2.5 Flare (22.30%).

Which top-10 model offers the lowest cost per task for image edit?

Gemini Nano Banana 2.1 balances #3 performance (score 1427) at $0.45 per completed task on Vigas.

How are Vigas leaderboard scores and win shares calculated?

Vigas combines real session outcomes across API workloads and agent harnesses with blind pairwise Arena votes using a Bradley-Terry statistical model with 95% confidence intervals.

Can I call the top image edit models through one API?

Yes. Point any OpenAI-compatible client at https://api.vigas.ai/v1 with your Vigas API key or set model to vigas/auto to route across providers automatically.