AI Model Directory Beta

Leonardo

Leonardo AI

Under evaluation

Maker: Leonardo
Origin: Australia
Released: 2023
Category: Image Generation

Strengths & weaknesses

Friendly all-rounder with its own fine-tuned models and style presets
Generous free daily allowance compared to most competitors

Free-tier limits push serious users toward a paid plan quickly
Less distinctive a look than Midjourney at the high end

Evaluation

Awaiting evaluation

Image qualityw40—
No independent measurement yet
Text renderingw20—
No independent measurement yet
Commercial designw20—
No independent measurement yet
Editingw20—
No independent measurement yet

w = weight, each criterion’s share of the overall score. Missing marks don’t count for or against. How scoring works

More in Image Generation

Midjourney2026.04.14

Midjourney V8.1

Under evaluation

Unmatched aesthetic taste and texture for concept art and exploration
Excellent for open-ended creative direction-finding

Weak at exact, repeatable text and structured brand templates
Discord/web-first workflow adds overhead versus API-first tools

OpenAI2026.04.21

GPT Image 2

8.7/10

Strong prompt-following for complex, multi-element compositions
Handles both generation and precise editing through one API

Small text, logos, and fine counts still need manual review
Less distinctive a look than Midjourney for pure aesthetic exploration

Google2025.11.20

Gemini 3 Pro Image

6.4/10

High-fidelity output up to 4K with strong reference-image handling
Good starting point for complex, detail-heavy briefs

Higher latency than Google's own Flash image tier
Text accuracy and identity preservation still need spot-checking

Black Forest Labs2025.11.25

FLUX.2

5.1/10

Flexible family of variants trading off quality, speed, and cost
Open-weight options available for self-hosted production pipelines

Confusing lineup; easy to pick the wrong variant for the job
Needs its own hosting and infra work to unlock full control