Grok Imagine Video 1.5 Gets Text-to-Video And Native 1080p

xAI has expanded the capabilities of Grok Imagine Video 1.5, the video generation model it released last month, adding several features that bring it closer to parity with other leading video generators. The headline addition is text-to-video support, meaning users can now generate a clip directly from a written prompt without needing to supply a reference image first. Previously, Grok Imagine's video output relied on an initial still image to anchor the generation. Alongside text-to-video, xAI has added voice references, a feature that lets a generated character maintain a consistent voice throughout a clip. This addresses one of the more noticeable weaknesses in AI video generation, where character voices can shift or lose consistency between cuts or scenes. Combined with image references, which allow users to guide the appearance of characters or settings using a supplied picture, the update gives creators more control over both the visual and audio identity of generated content. The model also now outputs video natively in 1080p, an upgrade from lower resolutions typically associated with earlier versions of the tool. Native high-definition output reduces the need for separate upscaling steps, which is likely to matter for users producing content intended for platforms like YouTube or social media where resolution and clarity affect viewer perception and platform recommendation algorithms. These updates position Grok Imagine Video 1.5 more competitively against other text-to-video tools on the market, several of which already offer text-only prompting and higher resolution output as standard. xAI has been moving quickly on its video generation roadmap since launching Grok Imagine, and this update suggests the company is treating video as a core part of its broader Grok product strategy rather than a secondary feature bolted onto its chatbot offering. Voice consistency in particular could prove useful for creators making narrative content, adverts, or explainer videos where a single character needs to speak across multiple shots without an obvious change in tone or pitch.