X AI Grok Imagine Video Reference To Video API Documentation

X AI Grok Imagine Video Reference To Video API Documentation

Playground

X-AI Grok Imagine Video Reference-to-Video generates videos from multiple reference images with preserved identity, style, and scene composition. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Grok Imagine Video Reference-to-Video is X-AI’s multi-image reference model that generates videos from up to 7 reference images. Provide reference images and describe the desired motion — the model generates a video that preserves the identity, style, and composition from your references with smooth, natural movement.

Why Choose This?

• Multi-image reference Use up to 7 reference images to guide video generation with rich visual context.

• Identity preservation Characters, objects, and scenes maintain consistent appearance across generated frames.

• Flexible duration Generate videos at 6 or 10 seconds to match your scene pacing.

• Resolution options Output in 720p or 480p based on your quality and speed requirements.

Parameters

How to Use

1. Upload your reference images — provide 1 to 7 reference images via URL or drag-and-drop upload.
2. Write your prompt — describe the motion, camera movement, and scene details. Reference the uploaded images in your prompt using @image1, @image2, etc.
3. Set duration — choose 6 or 10 seconds based on your scene length.
4. Select resolution — 720p for higher quality, 480p for faster processing.
5. Run — submit and download your video.

Pricing

Billing Rules

• Rate: $0.05 per second
• Duration options: 6 or 10 seconds
• Billing is based on the selected duration, not actual playback length

Best Use Cases

• Character Consistency — Generate videos with consistent character appearance across multiple shots using reference images.
• Product Showcases — Create dynamic product videos from multiple product photos.
• Multi-angle References — Use different angles of the same subject to generate richer, more accurate video.
• Social Media Content — Create engaging video clips from image collections for Reels, TikTok, and Shorts.
• Creative Projects — Combine multiple visual references to create unique video compositions.

Pro Tips

• Use high-quality, well-lit reference images for better identity preservation.
• Reference uploaded images in your prompt using @image1, @image2, etc. for precise control.
• Keep reference content and prompt aligned — if references show a character, describe that character’s actions.
• Start with fewer references and add more if needed for richer context.
• Use 6-second generations to test your prompt before committing to 10 seconds.

Notes

• Both images and prompt are required fields.
• Up to 7 reference images are supported.
• Ensure image URLs are publicly accessible.
• Maximum duration is 10 seconds.

Related Models

• Grok Imagine Video Image-to-Video — Generate video from a single image.
• Grok Imagine Video Extend — Extend existing videos with smooth continuation.
• Grok Imagine Video Edit — Edit existing videos with text instructions.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "images": [
    "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4"
  ],
  "duration": 6,
  "resolution": "720p"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/x-ai/grok-imagine-video/reference-to-video" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters