Back to all posts

Discover Top AI Image Tools With This New Leaderboard

2025-06-06Webb Wright3 minutes read
AI
Image Generation
Technology

The Rise of AI Image Generation

Before the widespread acclaim of ChatGPT, OpenAI made waves with DALL-E, its initial text-to-image AI model. Launched in January 2021, DALL-E offered an early glimpse into the transformative potential of AI-generated imagery. Fast forward to today, and the landscape is brimming with a diverse array of image-generating AI tools. Major players like OpenAI, Google, Midjourney, Stability AI, Ideogram, and Black Forest Labs each present their unique models, complete with distinct advantages and limitations.

(Disclosure: Ziff Davis, ZDNET's parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

If you're finding it challenging to keep up, you're not alone. For more insights, check out the best AI image generators of 2025.

gettyimages-1413205702-1 Image credit: gleitfrosch/Getty

Introducing LMArena ai Your Guide to AI Image Models

A new website, LMArena.ai, has emerged to help users navigate this evolving and increasingly packed field. Launched in beta last month, LMArena.ai ranks AI models based on their proficiency in generating images from text descriptions. This platform originated as a research initiative at UC Berkeley, initially known as Chatbot Arena, and has since grown into a comprehensive platform with backing from UC Berkeley, a16z, Sequoia Capital, and others.

If you're curious about practical applications, discover 3 image generators an AI tester actually uses.

How the Ranking Works

LMArena.ai employs the Elo rating system, famously used for ranking chess players, to determine the standings of different AI models. User votes are crucial to this process. To minimize bias, the names of the models are kept anonymous during the voting process; users only see the model names after they have cast their vote.

As of this writing, nearly 281,000 votes have been cast, evaluating the performance of nine text-to-image models. Currently, OpenAI's GPT Image 1 leads the pack, with Google's Imagen and Luma's Photon following closely.

The website also features charts below the scoreboard, offering a more detailed breakdown of how each model's ranking is calculated.

Cutting Through the Clutter in AI Imagery

The launch of ChatGPT in November 2022 ignited a surge in similar generative AI tools, as companies raced to establish dominance in what many predict could be a market as significant as smartphones or the internet itself. This rapid expansion has left users with a bewildering array of choices. Each generative AI model comes with its own set of capabilities, pricing models, and data privacy policies, making it a complex task to identify the best fit for individual needs. For instance, ChatGPT's own image generator has impressed many and is free to try.

Websites like LMArena.ai play a vital role in providing much-needed clarity. They are also beneficial for AI developers, offering platforms for valuable feedback derived from extensive testing. This continuous cycle of use and feedback contributes to the ongoing improvement of the AI models themselves. For a comprehensive comparison, ZDNET reviewed dozens of AI image generators, evaluating them on various factors like cost, safety features, speed, and quality.

Stay informed with top tech stories by subscribing to the Tech Today newsletter.

Read Original Post
ImaginePro newsletter

Subscribe to our newsletter!

Subscribe to our newsletter to get the latest news and designs.