Minimax H3 Highspeed Image To Video API Documentation

Minimax H3 Highspeed Image To Video API Documentation

Playground

MiniMax H3 Highspeed Image-to-Video generates fast 768P videos from input images with flexible 4-15 second duration, making it suitable for rapid image animation, social content, creative videos, ads, and production workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

MiniMax H3 Highspeed Image-to-Video turns a starting image into a video with prompt-guided motion. Provide a first-frame image, describe the action and camera movement, and optionally add a last-frame image to guide how the video ends.

Why Choose This?

• Image-to-video generation — Animate a still image with prompt-guided motion.
• Optional last frame — Add a target ending image for more control over the final transition.
• Flexible duration — Generate videos from 4 to 15 seconds.
• 768p output — Generate videos at 768p.
• Source framing preserved — Output aspect ratio follows the input image.

Parameters

How to Use

1. Upload a starting image — Provide the first frame for the generated video.
2. Write a prompt — Describe the desired motion, action, and camera movement.
3. Add a last image optional — Provide a target ending frame when additional control is needed.
4. Choose duration — Select a video length from 4 to 15 seconds.
5. Submit — Generate and retrieve the completed video.

Pricing

$0.06 per second of requested video duration.

The first-frame and optional last-frame images do not add separate charges.

Best Use Cases

• Portrait animation — Bring people and characters in still images into motion.
• Product videos — Animate product imagery for promotional content.
• Creative storytelling — Turn illustrations and concept art into moving scenes.
• Social content — Create short-form videos from existing images.
• Start-to-end transitions — Use a last-frame image to guide how the generated scene finishes.

Pro Tips

• Start with a clear image that has a strong main subject.
• Focus the prompt on motion, action, and camera behavior.
• Keep the requested action suitable for the selected duration.
• Use last_image when you want more control over the ending frame.
• Keep the first and last images visually compatible for smoother transitions.

Notes

• prompt and image are required.
• duration supports whole-second values from 4 to 15 and defaults to 5.
• Output resolution is 768p.
• Output aspect ratio follows the input image.

Related Models

• MiniMax H3 Highspeed Text-to-Video — Generate videos directly from text prompts.
• MiniMax H3 Highspeed Reference-to-Video — Generate videos using multimodal references.

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",
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "resolution": "768p",
  "duration": 5
}
JSON
)

# 1. Submit the prediction once. Never retry this request automatically:
# a retry after an unclear failure can create a second, billed task.
CHECK='The task may already exist. Check History or the predictions list before submitting again.'
if ! SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --connect-timeout 10 --max-time 60 \
  -X POST "https://api.wavespeed.ai/api/v3/minimax/h3-highspeed/image-to-video" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}"); then
  printf 'Submit failed. %s\n' "${CHECK}" >&2
  exit 1
fi

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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id. %s\n' "${CHECK}" >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction reaches a final status.
BODY_FILE=$(mktemp)
trap 'rm -f "${BODY_FILE}"' EXIT
while true; do
  # Transient errors (429, 5xx) are retried, honoring Retry-After.
  HTTP_CODE=$(curl --silent --show-error --fail-with-body \
    --connect-timeout 10 --max-time 60 --retry 5 --retry-delay 2 \
    --output "${BODY_FILE}" --write-out '%{http_code}' \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}" || true)
  case "${HTTP_CODE}" in
    2??) ;;
    000|429|5??)  # dropped connection or temporary error: keep polling
      printf 'Task %s: HTTP %s, polling again in 10s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      sleep 10; continue ;;
    *) printf 'Task %s: HTTP %s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      cat "${BODY_FILE}" >&2; exit 1 ;;
  esac
  RESULT=$(jq 'if type == "object" and has("data") then .data else . end' "${BODY_FILE}")
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted)
      printf 'Task %s ended with status %s\n' "${PREDICTION_ID}" "${STATUS}" >&2
      printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Every model also accepts the global request options enable_sync_mode and enable_base64_output; see their pages for details.

Response Parameters

Result Request Parameters

Result Response Parameters