xAI's Grok Imagine 1.5 adds full video generation and quality mode
Elon Musk touts Grok Imagine’s video gen at $0.07/sec – 4-megapixel output live now.
xAI has quietly transformed Grok Imagine from a basic image generator into a full-fledged video production platform with the release of Grok Imagine 1.5 Preview. Announced on June 3 and formally noted in xAI’s release notes on June 11, the update is now live — and Elon Musk directed his 200M+ followers to try it. The headline feature is full video generation, capable of producing clips up to 10–30 seconds at 4-megapixel native resolution using the Aurora-2 engine. Video is priced via API at $0.07 per second for 720p output (a 10-second clip runs ~$0.70), with a lower-tier 480p option at $0.05/second. The engine also integrates 'Voice Doctor' for context-aware audio synthesis that matches visuals, eliminating generic soundtracks.
Beyond video, Quality Mode — introduced around May 22 — delivers sharper textures, more accurate lighting, and dramatically improved text rendering compared to Speed Mode, solving a common AI art frustration. The 'Extend from Frame' feature, launched March 2, lets users chain 15-second clips for seamless longer sequences, ideal for storytelling or product demos. xAI also demonstrated a game asset prototyping workflow that chains image generation and image-to-video models, signaling professional creative use. The API, open since January 28, offers transparent pricing: $0.02 per text-to-image, $0.022 per image edit. With these rapid upgrades in under six months, Grok Imagine is no longer a consumer novelty but a serious tool for developers and creators.
- Video generation supports clips up to 30 seconds at 4-megapixel native resolution via Aurora-2 engine.
- Quality Mode (May 22) improves texture, lighting, and text rendering; API pricing at $0.07/sec for 720p video.
- Extend from Frame allows chaining 15-second clips; game asset prototyping pipeline demonstrated for professional use.
Why It Matters
xAI’s rapid six-month evolution makes Grok Imagine a viable competitor for professional video and asset creation workflows.