Wan 2.2 Animate 2

Wan 2.2 Animate 2

Playground

Wan 2.2 Animate 2 is the next-generation Wan character animation model: an end-to-end DiT that makes the character in a reference image perform the motion of a driving video, with no pose extraction, prompt-controlled background, and strong identity preservation; generates 480p/720p videos up to 120s. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

What is Wan 2.2 Animate 2?

Wan 2.2 Animate 2 (Wan2.2-Animate-2, the successor to Wan 2.2 Animate) is the next generation of the Wan character-animation family. It is an end-to-end model: the raw driving video conditions the generation directly inside a redesigned Diffusion Transformer, with no pose or keypoint extraction step. You provide one character image and one driving video, and the character performs the video’s full-body motion, hand gestures, and facial expressions.

Compared with Wan 2.2 Animate (v1):

• End-to-end motion transfer — no skeleton/pose preprocessing, which makes complex motion, fast dancing, and partially occluded performers noticeably more robust.
• Prompt-controlled background — the output background is generated from your prompt and is fully decoupled from both the reference image and the driving video. The background of your inputs does not matter.
• Strong identity preservation — the output character stays faithful to the reference image across the whole clip, including long videos.
• Long videos — driving videos up to 120 seconds are supported; segments are generated and chained natively by the model for seamless transitions.

The output video runs at 30 fps and follows the driving video’s duration and aspect ratio.

Note: Wan 2.2 Animate 2 focuses on character animation and does not include a replace mode. To swap a character into an existing shot while keeping the original scene and camera, use Wan 2.2 Animate (v1) in replace mode.

Pricing

Pricing depends on output length and resolution.

Billing rules

• Billed duration is clamped to the range 5–120 seconds

• If the video is shorter than 5 s → billed as 5 s

• If the video is longer than 120 s → billed as 120 s

• 480p: 0.20 USD per 5 s → 0.04 USD per second

• 720p: double 480p → 0.40 USD per 5 s → 0.08 USD per second

For example:

• 30 s at 480p → 30 × 0.04 = 1.20 USD
• 30 s at 720p → 30 × 0.08 = 2.40 USD
• 150 s at 720p (over the cap) → billed as 120 s → 120 × 0.08 = 9.60 USD

How to use Wan 2.2 Animate 2

1. image (required) Upload a clear character image. The character’s identity comes from here; the image background is ignored.

2. video (required) Upload the driving video whose motion and expressions you want to transfer. Up to 120 seconds.

3. prompt (optional) Describe how the character and the output background should look. The recommended format is:

Character appearance description: A young woman with shoulder-length pink hair, wearing a black leather jacket...
Background description: Plain light gray studio background, soft even lighting.

Describe looks only — put motion wording in motion_prompt.

4. motion_prompt (optional) A short description of the motion in the driving video, e.g. “A person doing energetic street dance with rhythmic steps”.

5. resolution Choose 480p for faster, cheaper previews or 720p for higher-quality output.

6. seed (optional) Leave default for a fresh variation each run, or set a fixed integer to reproduce a result.

Practical tips

• Match the framing — keep the reference image and the driving video similar in camera distance and crop (full-body to full-body, half-body to half-body). Framing mismatch is the most common cause of weak results.
• Describe the background in the prompt — since the background follows the prompt, an explicit background description gives you full control of the scene.
• Keep the performer clear — avoid heavy occlusion and strong motion blur in the driving video.
• Prototype at 480p — find good inputs with 480p and short clips first, then switch to 720p and longer durations.

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",
  "video": "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4",
  "resolution": "480p",
  "seed": -1
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.2/animate-2" \
  -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