On September 18, the official @grok account posted an invitation to “Make your own film scene with Grok @imagine,” alongside a video demo showing cinematic-quality clips generated from a prompt. Grok Imagine’s video capability has progressed since its July 2025 image-generation launch, through an October 2025 short-form video beta, the February 1, 2026 unveiling of Grok Imagine 1.0 with videos up to 10 seconds, 720p HD and synchronized audio, the June 16, 2026 release of Grok Imagine Video 1.5 via the xAI API, and a July 31, 2026 update adding native 1080p resolution, image references, voice references, and multi-reference scene control supporting up to seven simultaneous visual anchors per generation. The current version generates synchronized sound effects, ambience, and dialogue in a single inference pass, with maximum clip length at 15 seconds and a “Video Extend” feature pushing a sequence to 30 seconds total. It supports text-to-video, image-to-video, and prompt-based video editing. Grok Imagine Image 2.0 became generally available on August 7, 2026, focusing on precise prompt following, accurate text rendering within images, and multi-image references, with output at 1K or 2K resolution across various aspect ratios.