xAI, Elon Musk's artificial intelligence company, made Grok Imagine Video 1.5 generally available on June 17, 2026. The model takes a still image and a text prompt and produces clips up to 15 seconds long at 480p or 720p resolution, with audio generated in the same inference pass. Musk announced the wide release on X to all users that day.
On launch day, Grok Imagine Video 1.5 jumped 52 Elo points to claim the top position on Artificial Analysis's Image-to-Video Arena leaderboard, displacing ByteDance's Seedance 2.0 and Alibaba's HappyHorse-1.0. The faster Video 1.5 Fast variant produces a six-second 720p clip in roughly 25 seconds — down from over 40 seconds in its predecessor. The model is available at grok.com/imagine, via the xAI API under the identifier grok-imagine-video-1.5, and through xAI's iOS and Android apps. Pricing is set at $4.20 per minute.
The launch arrives during a pivotal shift in the AI video market. OpenAI deprecated Sora 2 on April 26, 2026, with the API slated for shutdown in September 2026, leaving a vacuum in premium video generation. The competitive field has since been shaped primarily by Chinese labs: ByteDance with Seedance 2.0 and Alibaba with HappyHorse-1.0 held the top leaderboard positions before xAI's entry. At $4.20 per minute, xAI's offering is 86 percent cheaper than Sora 2 Pro's $30-per-minute rate.
The model's core technical differentiator is single-pass audio-video synthesis — synchronized dialogue, ambient sound, sound effects, and background music are all generated without a separate audio pipeline. Lip-sync precision reportedly improved significantly over version 1.0, and cinematic camera movements such as dollies, pans, and tracking shots are described as more stable. Industry analysts note that native audio has shifted from a differentiator to a baseline expectation across top-tier video models; the benchmark competition in 2026 now centers on audio quality fidelity and motion consistency at longer clip durations.