AI Video Generator
Models
Sora Series

OpenAI's latest video generation model with enhanced quality

10 credits
10s/15s
Portrait/Landscape
Veo Series

Google's latest model with exceptional quality output

150 credits
Up to 8s
16:9/9:16
0/2000
10s
Public Visibility

When this option is enabled, the output video may be selected by AI Generate and published to the Explore.

Required Credits
30 6
80% OFF
Notification
This is a notification message
Sample Videos

Text and reference image video generation

Grok Imagine Video 1.5 generator

Build a short shot from a written scene or animate an existing image. Set the action, camera move, framing, duration, and resolution before generating, then review the result in the browser.

Grok Imagine Video 1.5 generator
01
1-15s
clip duration
02
6
aspect ratios
03
480p / 720p
output options
01

What Grok Imagine Video 1.5 is for

Grok Imagine Video 1.5 is best treated as a single-shot tool. It can turn a scene description into a clip or add motion to a supplied frame. The useful controls are concrete: who moves, what stays still, how the camera travels, where the action ends, and how long the shot lasts.

02

Controls that matter in a real workflow

Good results start with a shot plan, not a list of style words.

01

Text or reference input

Start from a written scene, or upload JPEG, PNG, or WebP references when the opening frame already contains the subject and composition you need.

02

Directed camera motion

Name one camera move such as a slow push-in, lateral track, locked tripod, tilt, or pull-back. Pair it with a clear end position.

03

Flexible framing

Use auto, square, landscape, portrait, 3:2, or 2:3. Choose the delivery format before writing blocking and subject position.

04

Draft-to-final resolution

Test movement and timing at 480p. Move to 720p after the action, framing, and source-image preservation are working.

03

Three practical ways to use it

Animate a product still

Keep the product geometry stable while adding controlled environmental motion.

  1. Upload a clean product frame
  2. Lock shape, label, color, and light direction in the prompt
  3. Add one camera move and one background action

Plan a cinematic insert

Write one beat that can finish cleanly inside a short clip.

  1. State the subject and starting position
  2. Describe the action in chronological order
  3. End with the final framing or pose

Build a longer sequence

Generate separate shots instead of forcing several scenes into one prompt.

  1. Break the idea into 3-8 second shots
  2. Repeat wardrobe, palette, and lens notes
  3. Join approved clips in an editor
04

Prompt templates you can adapt

Replace the bracketed details. Keep the action readable and give each shot one dominant camera instruction.

Reference-preserving product move

16:9

Use the uploaded image as the opening frame. Keep the [product] shape, label, materials, colors, and light direction unchanged. The camera makes a slow 15-degree orbit to the right while [background element] moves gently. End on a clean three-quarter product view. No extra objects or text.

Portrait motion

9:16

A close portrait of [subject] in [setting]. The subject turns toward the window, pauses, then looks back to camera. Hair and clothing move slightly in the breeze. Locked camera, shallow depth of field, stable facial features, continuous natural motion.

Atmospheric establishing shot

3:2

Wide establishing shot of [place] at [time]. A slow forward tracking move follows [subject] toward [landmark]. Foreground elements pass close to lens, distant movement remains subtle, and the shot ends with the landmark centered. One continuous take.

05

How to write a controllable video prompt

1

Write in shot order

Use starting frame, action, camera, environmental motion, and ending frame. Chronology is easier to evaluate than a paragraph of visual adjectives.

2

Limit competing motion

One subject action and one camera move are usually enough for a short clip. Too many independent actions make timing ambiguous.

3

Protect the source image

For reference work, list the identity, geometry, text, colors, camera angle, and lighting that must not change.

4

Test short and small first

Use a short 480p draft to check motion. Extend the duration or switch to 720p only after the shot direction works.

06

Supported settings on GhibliIA

  • Prompts up to 4,096 characters
  • JPEG, PNG, or WebP uploads, up to 5MB per file on this site
  • Durations from 1 to 15 seconds in one-second steps
  • Ratios: auto, 1:1, 16:9, 9:16, 3:2, and 2:3
  • Output resolutions: 480p and 720p
07

Grok Imagine Video 1.5 questions

Can I generate from both text and an image?

Yes. Text mode creates a scene from your prompt. Image mode sends uploaded reference URLs with the prompt so the source frame can guide the result.

How long can a clip be?

You can choose any whole-second duration from 1 to 15 seconds.

Which aspect ratios are supported?

The generator supports auto, 1:1, 16:9, 9:16, 3:2, and 2:3. When one reference image is used, the upstream API may follow the source image ratio.

How many credits does a video use?

480p costs 3 credits per second and 720p costs 5 credits per second. For example, an 8-second clip costs 24 credits at 480p or 40 credits at 720p.

How can I keep a face or product consistent?

Name every detail that must remain unchanged, then request a narrow action. Include identity, shape, printed text, light direction, camera angle, and color when they matter.