Elon Musk’s xAI announced a new AI-generated video model called Grok Imagine 1.0 on X (formerly Twitter), allowing users to create sophisticated AI videos. The updated video model is said to produce higher quality videos with improved audio output and is also cost-effective. The announcement of the new AI video model comes at an odd time, given that the platform has faced backlash over its AI image generation tools.
Grok Imagine 1.0 produces up to 10 seconds of AI video with 720p resolution and audio. This model allows users to transform text and static images into dynamic videos with movement and visual harmony.
For audio, we use multimodal AI to automatically generate and sync native audio. xAI says, “Experience a fully unfolding story. Sharper detail and smoother movement draw you deeper into the scene. Videos are now longer and in higher resolution.” In terms of performance, the company claims that the new video generation model has optimized latency, concurrency, and cost.
In addition to the Grok Imagine 1.0 model, the tech giant also released the Grok Imagine API, which combines video and audio generation and video editing capabilities. Therefore, users can seamlessly add, remove, and replace objects within the frame of a video. Users can also adjust scenes, colors, objects, etc.
According to xAI’s press notes, early benchmarks show the model outperforms competitors such as OpenAI’s Sora and Google’s Veo in processing speed and cost for short-form content.
xAI can integrate Grok Imagine into partner apps and tools. Developers also have direct access to Grok Imagine1.0 and APIs through xAI’s official APIs and SDKs.
This new product launch comes at a critical juncture for xAI, as xAI’s AI image generators have come under public and regulatory scrutiny for safety measures. Additionally, the company continues to be investigated by regulatory authorities around the world for producing non-consensual deepfakes.
