Breaking

Luma UNI-1.1 Max Lands 8th in Design Arena Image Ranking at Elo 1249

June 23, 2026 at 13:00 EDT

Media Generation
Foundation Models

Luma Labs AI's image-generation model "UNI-1.1 Max" has placed 8th (Elo 1249) on the Image Arena of the design-focused benchmark Design Arena, putting it in the same performance band as Google DeepMind's Gemini 3 Pro Image Gen 2K (codenamed Nano Banana Pro).

May 5, 2026 · Luma Labs AI

UNI-1.1 Max Cracks the Top 8 in Blind Image Showdown

Luma's "reasoning-first" image model lands 8th overall on Image Arena (Elo 1249) — matching Google's Nano Banana Pro band — while costing less than half as much per image.

8th

overall on Image Arena · Elo 1249

No.3

lab overall in image on Arena.ai

reference images supported per request

Price per image — Max vs comparable rivals

Luma says price and latency run less than half those of comparable models.

$0.10

Uni-1.1 Max

~2×+

comparable models

Standard Uni-1.1: ~$0.0404/image · extra reference: ~$0.003 each · 2048px default output

Where it ranks across Arena leaderboards

Graphic Design Arena

6th

1289

Logo category

6th

—

Image Arena (overall)

8th

1249

Reasoning-first pipeline

It plans intent, constraints, and composition before synthesizing pixels — aiming for prompt fidelity and brand consistency without heavy prompt engineering.

Plan intent & constraints → Compose layout → Synthesize pixels

What developers value

Strong cost-to-quality balance
Reasoning in iterative pro tasks — brand boards, sketch-to-photoreal, precise edits
Less prompt engineering, better consistency

Limitations noted

Occasional slips in highly complex scenes
Output depends on reference image quality
Rankings may shift with new votes & releases

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.