Hunyuan Video I2V API Documentation

Hunyuan Video I2V API Documentation

Playground

Hunyuan i2v turns images and text prompts into high-quality videos, generating coherent short clips from descriptive inputs. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Hunyuan Video I2V is an image-to-video model that turns a single reference image into a short animated clip guided by a text prompt. Upload an image to lock in subject, composition, and style, then describe the action and camera behavior you want. The model is well-suited for cinematic motion, character-driven beats, and atmospheric scenes where you want the “still” to come alive with coherent movement.

Key capabilities

• Image-to-video generation from one reference image
• Prompt-driven motion: actions, expressions, environment changes, camera movement
• Stable composition anchored to the input image
• Good for cinematic, dramatic, and stylized shots
• Supports duration and inference-step controls for quality vs. speed tradeoffs
• Multiple output sizes (e.g., 1280×720)

Use cases

• Animate key art, posters, and character portraits into short clips
• Cinematic micro-stories: close-up → reveal, slow push-ins, mood-heavy scenes
• Atmosphere and VFX-style motion: rain, fog, embers, neon flicker, drifting particles
• Social content loops from a single still image
• Rapid previsualization for scenes before full video production

Pricing

Inputs

• image (required): reference image to anchor subject and style
• prompt (required): action + camera directions

Parameters

• duration: clip length in seconds
• num_inference_steps: sampling steps (higher often improves coherence/detail)
• seed: random seed (-1 for random; set for reproducible results)
• size: output resolution (e.g., 1280×720)

Prompting guide (I2V)

Write prompts like a director’s brief:

• Subject: who/what is on screen
• Action: what changes over time (gestures, expression, environment)
• Camera: push-in, pull-back, pan, tilt, handheld vs. locked-off
• Mood/lighting: candlelight, moonlight, neon, fog, rim light
• Motion constraints: “subtle movement”, “no shaky camera”, “smooth dolly”

Example prompts

• An elderly lighthouse keeper in a wool sweater stands on a rocky cliff at dusk. He slowly turns his head toward the sea as the lighthouse beam sweeps across the darkening sky. Slow cinematic push-in, wind moving the grass, smooth motion, dramatic lighting.

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'
{
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "num_inference_steps": 30,
  "duration": 5,
  "size": "1280*720"
}
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/wavespeed-ai/hunyuan-video/i2v" \
  -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