Grok Imagine Video 1.5

AI video has been a leaderboard arms race for two years, and in June 2026 xAI took the top of it. Grok Imagine Video 1.5, announced on June 16-17, is an image-to-video model that ranks first on the Image-to-Video Arena, beating Sora 2, Google’s Veo 3.1, Seedance 2.0, and Kling in blind user comparisons. That alone would be newsworthy. What makes it genuinely disruptive is the price attached to that ranking: Grok Imagine Video 1.5 reaches the front of the pack at roughly 86% below the cost of the model it dethroned, which reframes the entire value conversation in the category.

The workflow is image-to-video at its core. You feed it a starting frame — a still image, a text prompt, or both — and it animates forward into a clip. The headline capability, and the one that separates it from much of the field, is native synchronized audio. Most video models generate silent footage you then have to score and sound-design separately; Grok Imagine Video 1.5 produces dialogue, sound effects, and background music in the same pass that produces the picture, all locked to the action on screen. For anyone making short social clips, ad concepts, or storyboards, collapsing video and audio into one generation is a real workflow saving, not a novelty.

Speed got a serious upgrade too. The model’s fast mode renders a 6-second 720p clip in about 25 seconds, down from more than 40 seconds in the previous version — quick enough to iterate on a shot in close to real time rather than waiting around between attempts. xAI also added sharper physics and more coherent motion, the two things that most often betray an AI clip as synthetic. It is shipping under a “preview” label, which is the right expectation to set: this is a fast-moving target where quality, limits, and moderation behaviour are all still being tuned in public.

Pricing is the story as much as the quality. Via the API, Grok Imagine Video 1.5 costs about $0.14 per second at 720p ($0.08 at 480p, $0.25 at 1080p), which works out to roughly $4.20 per minute at 720p — against Sora 2 Pro’s $30 per minute at its widescreen tier. Consumer access runs through grok.com/imagine, bundled into xAI’s Grok and X Premium subscriptions, where higher tiers unlock more generation headroom. That two-track model — cheap metered API for builders, subscription access for everyday users — is the same shape OpenAI and Google use, but xAI has undercut both on raw per-second cost by a wide margin.

Set against the video tools PixlRun already covers, the positioning is clarifying. Sora remains the name-brand benchmark and still has strengths in certain prompt styles, but it is now both beaten on the arena and dramatically more expensive. Veo by Google leans on its tight integration with Google’s ecosystem and strong prompt adherence; Kling and Runway Gen-3 have their own loyalists for specific looks and controls. Grok Imagine Video 1.5’s wedge against all of them is the combination of a leaderboard-topping result, built-in audio, and a price that makes high-volume generation actually affordable. For a creator or small studio generating dozens of clips a week, the cost difference alone can be the deciding factor before quality even enters the discussion.

The trade-offs are worth stating plainly. The first is lock-in: meaningful consumer access lives inside xAI’s subscription ecosystem, so getting the best of Grok Imagine means buying into Grok and, by extension, the X platform around it — a non-starter for anyone avoiding that ecosystem on principle. The second is the category’s standing limitation: like every model here, it is a short-clip generator. Six-to-fifteen-second shots with coherent motion are the sweet spot; long-form, narratively consistent video with stable characters across minutes is still beyond it, as it is for the entire field. And the “preview” status means today’s behaviour — including content moderation, which has been a recurring flashpoint for xAI’s image products — may shift under you.

The verdict: Grok Imagine Video 1.5 is the strongest price-to-quality proposition in AI video as of mid-2026, full stop. Topping the arena is impressive; topping it while costing a fraction of the incumbent and generating synchronized audio in the same pass is the kind of step-change that resets expectations for the whole category. The reasons to hesitate are ecosystem, not quality: if you are comfortable inside xAI’s world, this is close to a default recommendation for short-form video, and even Sora and Veo users owe it a serious look on cost alone. The short-clip ceiling and the preview-stage volatility keep it just shy of the top of our scale, but the trajectory is unmistakable.

One practical note on who this is for. Grok Imagine Video 1.5 is at its best in the hands of social-media creators, marketers building ad concepts, and designers storyboarding ideas who need a high volume of short, sound-complete clips and care about cost per clip. The image-to-video path — drop in a still you already have and animate it forward with matching audio — is especially well suited to turning existing product photography or key art into motion without a separate sound pass. It is less suited to filmmakers chasing minute-long continuity or brands that need airtight, predictable content moderation, where the preview-stage volatility is a genuine consideration. Used inside its sweet spot, though, the combination of a top arena ranking, native audio, and a per-second price this low makes it one of the most consequential releases the AI video space has seen this year, and the clearest sign yet that quality and affordability no longer have to be traded off against each other.