FrameSurfer publishes this guide and is one of the tools covered.
Grok Imagine is the image and video generator inside Grok, made by xAI, which now publishes as SpaceXSI. It has two video models: Grok Imagine Video, the original from January 2026, and Grok Imagine Video 1.5, from June 2026. The original is cheaper and stops at 720p. Video 1.5 adds native 1080p, better motion and sound, lip-synced speech and reference images.
On the xAI API, the original costs $0.05 per second and Video 1.5 costs $0.08 per second at 480p, as of October 13, 2026 (xAI models). In the Grok app, plans changed on October 12, 2026: Grok, X Premium and Cursor are now one subscription, from $8 a month.
Below: the two models side by side, a real clip, the new plans, the limits and what the API costs.
The two Grok Imagine video models
These details come from xAI's model pages and video guide, as of October 13, 2026.
| Grok Imagine Video | Grok Imagine Video 1.5 | |
|---|---|---|
| API name | grok-imagine-video | grok-imagine-video-1.5 |
| Released | January 28, 2026 | June 16, 2026, with references added July 31 |
| Resolution | 480p or 720p | 480p, 720p or native 1080p |
| Clip length | 1 to 15 seconds | 1 to 15 seconds (8 by default) |
| API price, 480p / 720p / 1080p | $0.05 / $0.07 / not offered | $0.08 / $0.14 / $0.25 per second |
| Input image | $0.002 each | $0.01 each |
| Sound | Made with the video | Made with the video, with lip-synced speech |
| Reference images | Yes, for clips up to 10 seconds | Yes, plus preset voices for up to 3 speakers |
| First and last frame, keyframes | No | Yes |
| Edit or extend a video | Yes | No |
Pick the original for cheap drafts, 720p social clips, and editing or extending a clip you already have. It is also the model xAI uses for its edit and extension tools.
Pick Video 1.5 when you need 1080p, when someone speaks on camera, or when a face or product has to stay the same across shots. You pay about 60% to 100% more per second for it.
xAI's API also lists a cheaper third option, Grok Imagine Video 1.5 Lite, at $0.02 per second at 480p, $0.03 at 720p and $0.14 at 1080p, as of October 13, 2026 (xAI docs). Its 1080p is upscaled from 720p, and xAI recommends it for simple videos.
What a Grok Imagine clip looks like
We ran this prompt on Grok Imagine Video, the original model, inside FrameSurfer on October 3, 2026:
A golden retriever puppy in a yellow raincoat splashes through puddles on a city sidewalk at dusk, low tracking shot, neon reflections on wet pavement, rain falling, cinematic.
It was one take at 720p, 16:9 and 5 seconds, with sound on. No retries and no editing. It took 60 seconds to render. The preview below is a short silent loop; the clip itself came with sound.
The puppy runs straight at the camera down the middle of a long, symmetrical street. The light came out brighter and more cheerful than the prompt asked for, closer to early evening than dusk. We ran the same prompt on five other models for our Sora alternatives test, if you want to compare.
How Grok Imagine video works
You describe a shot, or upload a still image and describe how it should move. Grok then makes the picture and the sound in one pass. xAI says Video 1.5 generates sound effects, ambience and dialogue together with the action.
One detail explains a lot. On Video 1.5, a text-only request runs in two steps behind the scenes. Grok first makes a still frame from your prompt, then animates it, according to the xAI video guide. You still send one request.
The models have changed quickly this year:
| Date | Release | What changed |
|---|---|---|
| January 28, 2026 | Grok Imagine API | Text to video, image to video and video editing in one API, as the grok-imagine-video model |
| June 16, 2026 | Grok Imagine Video 1.5 | Better motion, physics and audio. A faster "Video 1.5 Fast" version went into the Grok apps. |
| July 31, 2026 | Video 1.5 with References | Text to video without a start image, native 1080p, reference images and voice references |
| October 12, 2026 | New Grok plans | Grok, X Premium and Cursor sold as one subscription |
Reference images lock one thing in place, such as a face, a product or a location. You can then keep a character and change the scene, or keep the scene and change the action. xAI also says Video 1.5 Fast makes a 6-second 720p clip in about 25 seconds.
Grok plans after the October 12, 2026 change
On October 12, 2026, xAI replaced the SuperGrok plans with one subscription that covers Grok, Grok Bot, Cursor and X Premium (TeslaNorth, 24/7 Wall St.). Every paid tier draws on one shared AI usage pool. These are the US monthly prices reported at launch, with what the Grok plans page lists for each tier on October 13, 2026:
| Plan | US price | What the plan lists |
|---|---|---|
| Free | $0 | Limited Grok Chat and image generation |
| Plus | $8 a month | A shared usage pool for Grok Chat, Voice, Imagine and Build, plus the Grok API and Cursor |
| Premium | $30 a month | 4x the Plus usage, the latest Grok models, Grok Bot, and "Make stunning AI images & videos" |
| Super | $100 a month | 15x the Plus usage and priority model access at peak times |
| Ultra | $200 a month | 50x the Plus usage and the earliest access to new products and models |
A few things to know:
- Launch offers. On October 13, the plans page offered 25% off Plus for six months and 50% off the first month of Premium, Super and Ultra.
- Local prices. The page shows prices in your currency. From Europe it showed €7, €31, €105 and €210 a month.
- Old plans. The company says nothing changes for people already on SuperGrok, Cursor or X Premium. Switching to a new plan cancels the old one with a prorated refund.
Is Grok Imagine free?
Not for video, based on the new plan list. The Free tier lists limited chat and image generation. Plus includes Imagine in its shared usage pool, and Premium is the first tier whose list says "Make stunning AI images & videos". If a site offers free Grok Imagine video, check who runs it before you sign up.
Grok Imagine video limits
The app and the API work differently, so the limits differ too.
| Grok app (web, iOS, Android) | xAI API | |
|---|---|---|
| Clip length | Not listed per plan since October 12 | 1 to 15 seconds per request |
| Resolution | Not listed per plan since October 12 | 480p (the default) or 720p, and 1080p on Video 1.5 |
| Starting points | Text, a photo, reference images and voices | Text, an image, reference images, or a video to edit |
| Sound | Made with the video | On by default, and you can ask for a silent clip |
| How you pay | A shared usage pool | Per second of video |
Video uses far more of the shared pool than a chat message, so heavy video work fills it fastest. And a few API limits:
- Aspect ratios. 1:1, 16:9, 9:16, 4:3, 3:4, 3:2 and 2:3, with 16:9 as the default. For image to video, the clip follows your image's shape. If you set a different ratio, the image gets stretched.
- Editing. You can edit a source video of up to 8.7 seconds on the original model. The result keeps that length and is capped at 720p.
- References. Reference-to-video output is capped at 720p, and at 10 seconds on the original model.
Grok Imagine API pricing (as of October 13, 2026)
The API bills per second of finished video, by model and resolution. These rates come from xAI's model pages for grok-imagine-video, grok-imagine-video-1.5 and grok-imagine-video-1.5-lite.
| API model | 480p | 720p | 1080p | Each input image |
|---|---|---|---|---|
| grok-imagine-video | $0.05 per second | $0.07 per second | Not offered | $0.002 |
| grok-imagine-video-1.5 | $0.08 per second | $0.14 per second | $0.25 per second | $0.01 |
| grok-imagine-video-1.5-lite | $0.02 per second | $0.03 per second | $0.14 per second | $0.01 |
What that means for one 10-second clip:
| Clip | grok-imagine-video | grok-imagine-video-1.5 | grok-imagine-video-1.5-lite |
|---|---|---|---|
| 10 seconds at 480p | $0.50 | $0.80 | $0.20 |
| 10 seconds at 720p | $0.70 | $1.40 | $0.30 |
| 10 seconds at 1080p | Not offered | $2.50 | $1.40 |
Other pricing notes from the same pages:
- Text input is free.
- On
grok-imagine-video, a video you send in as input costs $0.01 per second. - On
grok-imagine-video-1.5, preset voice audio input is free. - xAI says video extension pricing is a promotional rate and may change.
Watermarks, moderation and commercial use
These rules come from xAI's pages, as of September 2026.
- Watermark. Images and videos made in the Grok website and apps carry a Grok watermark that marks them as AI-made. There is no setting to remove it. The Grok FAQ says removing or hiding it breaks xAI's Acceptable Use Policy.
- Moderation. The Imagine API page says all generated images and videos go through automatic content moderation. Content that breaks xAI's usage policies may be blocked. API results also include a moderation flag.
- Ownership. The Grok Imagine page says what you create is yours, including for commercial use.
Tips for better Grok Imagine clips
These follow from how the models work:
- For image to video, describe the change. The photo already sets the look. Spend your words on motion, camera and sound.
- Pick the shape first. Choose 9:16 for Reels and TikTok before you generate, or start from an image that already has the right shape.
- Plan in shots. Each API request tops out at 15 seconds, so write a longer video as a list of clips.
- Use Video 1.5 for speech and continuity. Lip-synced dialogue and reference images are where 1.5 earns its higher price.
- Draft on the cheaper model. Test the idea on the original or 1.5 Lite, then re-run the keeper on Video 1.5.
Grok Imagine in the app, the API or FrameSurfer
| Route | Best for | Trade-offs |
|---|---|---|
| Grok app | Quick clips, references and Video 1.5 | Video shares one usage pool with chat, coding and X. Every clip carries a watermark. |
| xAI API | Developers who want per-second billing and a choice of all three models | You build the editing, captions and voiceover yourself. |
| FrameSurfer | Finished short videos | Runs the original Grok Imagine Video at 720p, 1 to 15 seconds. Video 1.5 and references are not offered. |
On FrameSurfer, Grok Imagine Video makes clips of 1 to 15 seconds at 720p. It starts from text or a start image and generates native audio. You can set the aspect ratio, including when you start from an image, plus the resolution and duration. It sits next to Kling 3.0, Veo 3.1 Fast, Seedance 2.5, Wan 3, MiniMax H3 Max Turbo and Pruna P-Video-2 on the same plan. Pick Grok for sound-on shots at 720p, and another model where you need 1080p or a longer clip.
FrameSurfer also does the parts a single clip leaves out. Give it an idea, a photo, a script or a product link, and it writes the script, generates every shot, and adds voiceover, captions and music. You refine the result in a visual editor and export an MP4 at up to 1080p for TikTok, Reels, Shorts or ads.
To try Grok Imagine on one shot, open FrameSurfer's text to video tool and pick the model.
How Grok Imagine compares with other models
- MiniMax H3, the model behind Hailuo AI, adds 2K output and mixed image, video and audio references. See the Hailuo AI video guide.
- Alibaba Wan 3 makes single clips of up to 30 seconds. See the Wan AI video generator guide.
- Kling 3.0 adds multi-shot scenes and dialogue, and Kling 4.0 is in early access. See Kling AI pricing.
- Sora 2 users looking for a replacement can start with Sora alternatives.
- For every model FrameSurfer runs, side by side, see AI video models compared. For whole tools, see the best AI video generators.
Starting from a product shot or a portrait? FrameSurfer's image to video tool makes your photo the first frame and animates it with Grok Imagine or any other model on your plan.
Frequently Asked Questions
What is the difference between Grok Imagine Video and Video 1.5?
Grok Imagine Video is the original model from January 2026. It is cheaper, stops at 720p, and handles video editing and extension. Video 1.5, from June 2026, adds native 1080p, lip-synced speech, reference images and first and last frame control. On the API it costs $0.08 to $0.25 per second, against $0.05 to $0.07 for the original.
Is Grok Imagine video free?
Not on xAI's own plans. Since October 12, 2026, the Free tier lists limited chat and image generation. Plus ($8 a month) includes Imagine in a shared usage pool, and Premium ($30 a month) is the first tier that lists AI videos.
How much does Grok cost now?
Since October 12, 2026, Grok, X Premium and Cursor are one subscription: Plus $8, Premium $30, Super $100 and Ultra $200 a month in the US, as reported at launch. Prices on grok.com show in your local currency, and launch discounts apply to the first months.
How long can Grok Imagine videos be?
Each API request makes a clip of 1 to 15 seconds on both models. Reference-to-video on the original model stops at 10 seconds. The new app plans do not list a clip length per tier.
Does Grok Imagine video have sound?
Yes. Clips include an audio track by default, with sound effects, ambience and dialogue generated with the picture. Video 1.5 adds lip-synced speech. The API can also return a silent clip if you ask for one.
How much does the Grok Imagine API cost?
As of October 13, 2026, grok-imagine-video costs $0.05 per second at 480p and $0.07 at 720p. grok-imagine-video-1.5 costs $0.08, $0.14 and $0.25 per second at 480p, 720p and 1080p. grok-imagine-video-1.5-lite costs $0.02, $0.03 and $0.14.
Can I remove the Grok watermark?
No. xAI's FAQ says videos made in the Grok website and apps carry a Grok watermark, with no setting to remove it. Removing it breaks xAI's Acceptable Use Policy.
Do I need X Premium to use Grok Imagine?
Not separately anymore. Since October 12, 2026, the paid Grok plans include X Premium benefits, so one subscription covers both.
