Pruna AI P Video Avatar API Documentation

Pruna AI P Video Avatar API Documentation

Playground

Pruna AI P-Video Avatar is a fast AI avatar video generation model that creates high-quality avatar videos for digital humans, talking characters, social media content, marketing creatives, virtual presenters, and AI video workflows. Ready-to-use REST inference API with simple integration, no coldstarts, and affordable pricing.

Features

Pruna AI P-Video Avatar generates a talking or performing avatar video from a reference image and an audio clip, with optional prompt guidance for motion and expression. It is designed for character-driven video generation where the image defines the avatar and the audio drives the timing and delivery.

Why Choose This?

• Image + audio avatar generation Combine a reference image with an audio track to generate a video avatar performance.

• Prompt-guided motion control Use video_prompt to nudge expression, movement, or overall performance style.

• Simple output settings Choose resolution and seed without a heavy configuration workflow.

• Audio-driven timing Video length follows the uploaded audio duration, making it easier to generate synced outputs.

• Production-ready workflow Useful for avatar clips, talking portraits, character presentations, and short-form content generation.

Parameters

How to Use

1. Upload your image — provide the reference image you want to animate.
2. Upload your audio — use a clear audio clip to drive the avatar performance.
3. Add a simple video prompt (optional) — describe only the key motion or mood you want.
4. Choose resolution — use 720p for lower cost or 1080p for higher quality.
5. Set a seed (optional) — use a fixed seed for more reproducible outputs.
6. Submit — run the model and download the generated avatar video.

Example Prompt

Natural head movement, subtle facial expression, stable identity, clean speaking performance, realistic motion

Pricing

> Billing duration: Audio duration is rounded up to whole seconds and capped at 600 seconds. No minimum-duration floor is applied.

Pricing is based on the audio duration and resolution.

720p

1080p

Billing Rules

• Pricing is based on the uploaded audio duration
• Maximum billed duration is 600 seconds
• 720p uses a base rate of $0.025 per second
• 1080p costs 1.8× the 720p rate
• video_prompt and seed do not affect pricing

Best Use Cases

• Talking avatar videos — Generate speaking portraits from a single image and audio track.
• Character presentation clips — Create short performance-based videos for storytelling or demos.
• Social media avatar content — Produce short avatar-driven clips for lightweight content workflows.
• Narration-driven character scenes — Pair a static character image with voice content for expressive video output.
• Prototype virtual presenters — Quickly test avatar-based presentation ideas without full animation workflows.

Pro Tips

• Keep the audio reasonably short for better reliability and easier iteration.
• Use a clear, front-facing image for better avatar stability.
• Keep video_prompt simple and direct — overly detailed prompts are more likely to fail.
• Focus the prompt on a few essentials, such as natural motion, subtle expression, or stable identity.
• Start with 720p for testing, then switch to 1080p for final-quality outputs.
• Reuse the same seed when you want more consistent variations.

Notes

• Both image and audio are required.
• Very long audio is not recommended; shorter clips are easier to generate successfully.
• The model works best when video_prompt is simple rather than highly detailed.
• Billing uses the rounded-up audio duration with no minimum-duration floor and a cap of 600 seconds.
• save_audio is not exposed in the current input settings shown here.

Related Models

• Pruna AI P-Video Text-to-Video — Generate videos directly from natural-language prompts.
• Pruna AI P-Video Image-to-Video — Animate a reference image into a video clip with prompt guidance.
• Pruna AI P-Image Text-to-Image — Generate still images for image-first creative workflows.

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",
  "video_prompt": "The person is talking.",
  "resolution": "720p"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/pruna-ai/p-video/avatar" \
  -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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 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|deleted) 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