Grok Imagine is xAI's image and video generator inside the Grok and X apps. Clips made there can carry a visible Grok mark and are metered by subscription tier, with caps that xAI has reshuffled several times in 2026. The same xAI models, Grok Video and Grok Imagine Video 1.5, run on Aura AI with no watermark, at 720p, up to 15 seconds with native audio, priced per generation in credits on every plan. This guide covers what the app applies, what the tiers allow, and the exact specs and credits on Aura AI.
In This Article
What Grok Imagine is
Grok Imagine is the image and video generator built into the Grok app and into Grok on X. You type a prompt or start from a picture, and it returns stills or short clips with sound; the video side became the headline feature through 2025 and 2026 because it animates an image quickly and adds audio in the same pass. The models behind it are xAI's Grok Imagine Video line, now at version 1.5, which xAI released on May 31, 2026 and which currently tops the Image-to-Video Arena leaderboard.
Two things frustrate people who use it inside the app: the mark on the output and the meter on usage. Both come from the app, not from the model.
The Grok Imagine watermark
Clips and images generated in the Grok app can carry a visible Grok identifier on the output. The Grok model page on Aura AI puts it plainly: xAI's native app may include identifiers on output, while the same models on Aura AI return a clean file. Like every burned-in mark, it reads as a repost when it lands in an ad or a client edit, and cropping or blurring it costs framing and sharpness on a clip that is already 720p or lower.
Watermark versus provenance
A visible identifier is a branding choice by the app. Invisible provenance metadata is a separate, industry-wide layer that many generators attach and that you should leave alone. The first is the one that matters for professional use; the second does not affect quality.
Daily limits and tiers in the Grok app
xAI has reshuffled Grok Imagine's caps repeatedly in 2026, so treat every figure as a snapshot. Based on third-party trackers and support emails quoted by users in May and June 2026:
| Grok app tier | Reported price | Reported video access (mid-2026) | Watermark |
|---|---|---|---|
| Basic (no subscription) | $0 | Image and video generation reported as reserved for paid tiers, or heavily rationed | Yes |
| SuperGrok Lite | About $10/mo | Entry point for video; reported at roughly 3 videos a day, 480p, 6-second clips | Yes |
| SuperGrok | About $30/mo | Reported at more than 20 videos per 24 hours after a May 2026 revision | May apply |
| SuperGrok Heavy | Higher tier | Reported at more than 80 videos per 12 hours; from June 2026 usage moved to a shared weekly pool across chat and media | May apply |
The June 2026 change is the one that catches people out: instead of a daily video count, Grok now draws chat and media from one weekly quota, and a single video can consume a noticeable slice of it. If you generate in bursts, the cap arrives mid-project. Numbers above are as reported and may have changed again by the time you read this.
Grok Video and Grok Video 1.5 on Aura AI: no watermark, priced per clip
The same xAI models run on Aura AI without the app around them. Output is delivered with no watermark, there is no daily or weekly video cap, and each generation is priced in credits that you see before you run it. Specs below are from the Grok model page, updated September 2026.
| Model on Aura AI | Mode | Credits | Resolution | Duration | Audio | Plan |
|---|---|---|---|---|---|---|
| Grok Imagine Video 1.5 | Image to video | 4 credits for 5 s at 720p, +0.4 per extra second; 480p x0.7 | 480p or 720p | Up to 15 s | Native, same pass | Every plan |
| Grok Video | Text to video, image to video | 3 credits for 5 s, +0.3 per extra second; 15 s at 720p is 6 credits | 480p or 720p | 3 to 15 s | Native | Every plan |
| Grok Reference | Reference to video, up to 7 images on Premium | 4 credits, +0.5 per second at 720p | 720p | 3 to 10 s | Native | Every plan (reference count by plan) |
| Grok Image / Grok 2.0 Image | Text to image | 1 credit per image; Grok Edit 2 credits | High resolution | Still | n/a | Every plan |
Seven aspect ratios are available on Grok Imagine Video 1.5, from 16:9 to 9:16, and most clips finish in well under a minute. A 15-second Grok Video clip at 720p costs 6 credits; the same length on Grok Imagine Video 1.5 at 720p is 8 credits. Video 1.5 ships as image-to-video first, so start from a still: generate one with Grok Image at 1 credit, or upload a product photo, then animate it through image to video. Text prompts with no source image go to the base Grok Video model in text to video.
Grok app versus Aura AI, side by side
| Grok Imagine in the X / Grok app | Grok Video 1.5 on Aura AI | |
|---|---|---|
| Watermark | Visible identifier may be applied | None |
| Usage cap | Daily counts, then a shared weekly pool from June 2026 | Credits per generation, monthly allowance by plan |
| Resolution | 480p on the entry tier, 720p above | 720p, with 480p as a cheaper option |
| Duration | 6 s on the entry tier, longer above | Up to 15 s |
| Audio | Yes | Native, same pass |
| Other models | Grok only | Kling 3.0, Seedance 2.5, Veo 3.1, Sora 2 and 40+ more on one balance |
| Commercial use | Per xAI terms | Cleared, no watermark |
When Grok is the right pick, and when it is not
Grok Imagine Video 1.5 is the model to reach for when you have a still and want it moving with sound in seconds at a low credit cost: product shots, meme edits, social clips, quick animatics. Its ceiling is 720p and 15 seconds. When the brief needs 1080p or 4K, precise lip sync or a longer take, switch models on the same balance: Kling 3.0 runs 3 to 15 seconds at 1080p from 3 credits with the most accurate mouth movement, and Seedance 2.5 generates picture and speech in one pass at up to 1080p with up to 15 seconds and six aspect ratios. All three export with no watermark, so the decision is about resolution, length and look rather than about cleaning up the file afterwards.
How to do it in Aura AI
- Open image to video in the web app and upload the still you want to animate, or generate one first with Grok Image at 1 credit.
- Select Grok Imagine Video 1.5 in the model selector, pick 480p or 720p, the aspect ratio and a duration up to 15 seconds. The credit cost updates as you change settings.
- Describe the motion and the sound. Say what moves, how the camera behaves and what you want to hear; Grok reads audio cues from the prompt.
- Generate and download the MP4 with no watermark. For a text-only prompt, use text to video with the base Grok Video model instead.
Run Grok Imagine Video 1.5 with no watermark
xAI's latest video model at 720p, up to 15 seconds with native audio, from 4 credits per clip, on every plan.
Open Image to VideoFrequently Asked Questions
Does Grok Imagine put a watermark on videos?
Output generated inside the Grok and X apps can carry a visible Grok identifier. Grok Video and Grok Imagine Video 1.5 generated through Aura AI are delivered with no watermark and are cleared for commercial use.
What are the Grok Imagine daily limits?
They change often. In mid-2026, third-party trackers reported roughly 3 videos a day at 480p and 6 seconds on SuperGrok Lite, more than 20 videos per 24 hours on SuperGrok, and more than 80 per 12 hours on SuperGrok Heavy, before xAI moved usage to a shared weekly pool in June 2026. On Aura AI there is no daily cap; each clip costs credits from your monthly allowance.
What does Grok Imagine Video 1.5 cost on Aura AI?
4 credits for a 5-second clip at 720p, plus 0.4 credits per extra second, with 480p at 0.7 times the price. A 15-second 720p clip is 8 credits. The base Grok Video model is 3 credits for 5 seconds plus 0.3 per extra second, so 15 seconds at 720p is 6 credits. Both are available on every plan.
Does Grok Video 1.5 generate audio?
Yes. Grok Imagine Video 1.5 synthesizes native audio in the same pass as the picture at 720p, with clips up to 15 seconds. No separate audio step is needed.
Can I use Grok Video from a text prompt only?
Yes. Grok Imagine Video 1.5 on Aura AI is image-to-video first, so start from a still or generate one with Grok Image at 1 credit. For a text-only prompt, pick the base Grok Video model in the text to video tool; it supports 3 to 15 seconds at 480p or 720p with native audio.