Pruna AI P Video 2 Pro Avatar API Documentation

Pruna AI P Video 2 Pro Avatar API Documentation

Playground

Pruna P-Video-2-Pro Avatar creates talking-avatar videos from a portrait image and uploaded speech audio, making it suitable for digital humans, presenters, dubbing, and avatar video workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

P-Video-2 Pro Avatar creates a talking-avatar video from a portrait image and a speech audio recording. The portrait defines the character and visual appearance, while the supplied audio drives the speaking performance and determines the video duration.

Use an optional prompt to guide expression, movement, or framing, and choose between 480p and 768p output for avatar presentations, explainers, social content, and other speech-driven video workflows.

Why Choose This?

• Image-and-audio avatar generation
  Turn a portrait image and recorded speech into a talking video.

• Audio-driven performance
  The input audio controls the speaking performance and output duration.

• Up to 3 minutes of speech
  Generate avatar videos from audio clips up to 180 seconds.

• Prompt-guided movement
  Describe desired expressions, gestures, movement, or framing through an optional prompt.

• 480p and 768p output
  Choose the resolution that fits your quality and cost requirements.

• Seed control
  Use an optional seed when reproducibility is useful.

Parameters

How to Use

1. Upload a portrait — Provide a clear image with the person’s face visible.
2. Upload speech audio — Supply a clean speech recording.
3. Add performance guidance optional — Use prompt to describe expressions, gestures, movement, or framing.
4. Choose resolution — Select 480p or 768p.
5. Set seed optional — Use a fixed seed when reproducibility is needed.
6. Submit — Generate and retrieve the talking-avatar video.

Pricing

Pricing is based on the input audio duration and selected resolution.

Audio duration is rounded up to the next whole second, with a minimum of 1 billed second and a maximum of 180.

prompt, prompt_upsampling, and seed do not add separate charges.

Best Use Cases

• Virtual presenters — Create avatar-led introductions, announcements, and presentations.
• Product explainers — Turn recorded narration into a character-led product presentation.
• Educational content — Generate presenter-style videos for lessons, training, and instructional material.
• Social media content — Create talking-character videos from existing portraits and speech.
• Digital characters — Animate character portraits with recorded dialogue.
• Localized presentations — Reuse the same portrait with different speech recordings for multiple versions.

Pro Tips

• Use a clear, well-lit portrait with the face fully visible.
• Use clean speech audio without loud background music or overlapping speakers.
• Keep movement instructions simple and consistent with the source portrait.
• Describe natural expressions and gestures when you want more than a basic talking performance.
• Use 480p for lower-cost testing before moving to 768p.
• Start with shorter audio clips when testing a new portrait or prompt.
• Use a fixed seed when comparing prompt changes.

Notes

• image and audio are required.
• Use an audio clip from 1 to 180 seconds, up to 64 MiB.
• Portraits support JPG, JPEG, PNG, and WebP, up to 20 MiB.
• prompt_upsampling is off by default.
• Output video duration follows the processed input audio.
• Billing duration is rounded up to the next whole second.
• The minimum billed duration is 1 second and the maximum is 180 seconds.
• resolution defaults to 768p.
• prompt is optional and defaults to The person is talking.
• Public URLs must remain accessible while the request is processing.

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",
  "audio": "https://interactive-examples.mdn.mozilla.net/media/cc0-audio/t-rex-roar.mp3",
  "resolution": "768p",
  "prompt_upsampling": false
}
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/pruna-ai/p-video-2-pro/avatar" \
  -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

Response Parameters

Result Request Parameters

Result Response Parameters