Breaking

xAI's Grok Imagine Video 1.5 retakes I2V top spot at Elo 1357

June 14, 2026 at 02:30 EDT

Media Generation
Foundation Models

On June 3, 2026, xAI officially unveiled "grok-imagine-video-1.5-preview," a preview model specialized in Image-to-Video generation, which retook the top spot on the blind-evaluation platform Design Arena's Image to Video leaderboard with an Elo of 1357.

June 3 · xAI

Grok Imagine Video 1.5 takes #1 on the Image-to-Video leaderboard

xAI's new preview model animates a single still image with prompted motion, camera work and physics — and embeds native audio with lip sync at generation time. It tops Design Arena's crowdsourced board with a 1357 Elo.

1357

Elo · Image-to-Video
1473 reported for 720p

41.2s

Avg generation time
at 480p

720p

Max resolution
24 fps

Audio

Native sound + lip sync
at generation time

Pricing — pitched on a speed / cost frontier

Per-second cost scales by output resolution; image input is near-flat.

$0.01

Image
input

$0.08

480p
per sec

$0.14

720p
per sec

720p costs 1.75× more per second than 480p. xAI pitches lower cost and shorter latency against higher-priced rivals like Veo 3 and Seedance 2.0.

Strengths

Strong motion, lighting and cinematic texture
Native audio + lip sync — a portrait gets synced lines
Can edge out Seedance 2.0 with camera-work techniques
Fits short-form, explainers, product ads, anime shorts

Limitations

API-only; uneven rollout to consumer apps
No direct text-to-video — still image input required
Complex motion strains temporal coherence
Outputs center on short clips of a few seconds to 15s

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.