Grok Imagine Image 2.0 (Low) Debuts at #2 on Text-to-Image Arena
Grok Imagine Image 2.0 (Low) has debuted at #2 on both the Text-to-Image Arena with 1320 points and the Image Edit Arena with 1439 points. This represents a major leap from its predecessor, which previously ranked #14 and #7 in those respective categories. This release highlights a rapid improvement in xAI's image generation capabilities, positioning its model as a top competitor against other leading generative AI systems. It demonstrates that Grok's underlying image models are becoming highly competitive in human-preference benchmarks. The new model is currently only accessible through the official app and is not available via an API. Additionally, it secured the #2 spot across all evaluation categories except for Art, which is still awaiting more votes to finalize its score.
## BACKGROUND
The Text-to-Image Arena is a public benchmarking platform where users compare side-by-side image outputs from different AI models to rank them using Elo rating systems. Grok Imagine is the image generation feature developed by Elon Musk's AI company, xAI, integrated into their Grok assistant.