Eseandre Mordi
xAI's Grok Imagine has climbed to third place on the Video Editing Arena leaderboard, marking another milestone in the rapid evolution of AI-powered video systems. The 480p version of the model now holds the #1 overall win rate at 66%, outperforming established competitors while sitting just three Elo points behind the second-ranked system. With a 720p version already in development, Grok Imagine's trajectory points to an increasingly competitive landscape where marginal improvements in resolution and editing fidelity may decide future rankings.
Grok Imagine Achieves 66% Win Rate and 1253 Elo Score
Grok Imagine has secured a Top 3 position on the Video Editing Arena leaderboard, as shown in the DesignArena image. The model achieved the #1 overall win rate of 66%, ahead of Kling O1 Edit at 65% and Gen-4 Aleph at 63%. In the Elo category, Grok Imagine ranks third with 1253, closely trailing Gen-4 Aleph at 1256 and Kling O1 Edit at 1278. The tweet emphasizes that this performance reflects the 480p version of the model, underscoring the rapid progress of xAI's video systems.
The three-point gap between Grok Imagine and Gen-4 Aleph indicates near parity in competitive strength under the Elo metric. While Kling O1 Edit leads in Elo rating, Grok Imagine currently holds the strongest win rate across evaluated matchups.
This performance reflects the 480p version of the model, underscoring the rapid progress of xAI's video systems.
This convergence of scores highlights accelerating competition in AI-driven video editing models. Broader advancements in the space have also been noted in ByteDance's 5-billion-parameter OpenVE3M model sets new bar for AI video editing.
720p Version Expected to Push Rankings Higher
According to the announcement, a 720p version of Grok Imagine is expected to deliver even stronger results once evaluated. The model's rapid improvement trajectory aligns with other reported milestones, including xAI Grok Imagine hits #1 spot in global AI video rankings with 1397 score. The ongoing leaderboard shifts suggest continuous optimization and iteration across competing labs.
As video generation and editing models advance, ranking systems such as DesignArena provide comparative insight into performance under real evaluation conditions. Competitive movements across categories have also been observed in adjacent AI leaderboards, including Kimi 2.5 tops DesignArena rankings with 8-category lead over Gemini and Claude. The tightening margins across Elo scores indicate that incremental improvements in resolution, stability, and generation fidelity may determine future leaderboard leadership in AI video editing.
Eseandre Mordi
