Skip to content

Pruna AI

P-Video-2-Pro Avatar

Animate a portrait using a recorded voice or a written script. The model matches mouth movement and facial expression to the speech and supports videos up to three minutes.

API pricing · From $0.02/output sec
API model IDpruna/p-video-2-pro-avatar
One API keyA studio presenter speaks the supplied ElevenLabs recording, listing four media types and explaining how to switch models.1152 × 768 · 24 fps

Best for

Technical details

Read the model documentation
API model ID
pruna/p-video-2-pro-avatar
Provider
Pruna AI
Modality
Video
Input
Text prompt · Images · Voice script · Audio
Output
Video
Endpoint
POST /v2/jobs
Workflow
Avatar video
Resolution
480p · 768p

Request parameters

  • imagestring · Required

    Public https URL of the portrait: jpg, jpeg, png or webp, up to 20 MiB. It is the first frame and sets the aspect ratio.

  • promptstring

    What the video should show. Default "The person is talking."

  • resolutionstring

    Output resolution, each with its own per-second rate. Default 768p. The higher tiers, 1080p and above, are not offered yet.

    • 480p
    • 768p
  • audiostring

    Public https URL of speech audio, 1 second to 3 minutes and up to 64 MiB. It sets the video's length and wins over voice_script. Send this or voice_script.

  • voicestring

    Voice for voice_script: one of 30 named voices, each written with (Female) or (Male) after the name. Default Zephyr (Female).

Additional parameters

  • enable_safety_filterboolean

    Turns the generator's content check on or off. Default false.

  • prompt_upsamplingboolean

    Adds the scene's own motion, read from the portrait, after your words; false uses the prompt exactly as written. Default true.

  • seedinteger

    Fixes the random draw: the same inputs and seed give the same result.

  • voice_languagestring

    Language of voice_script: English (US) (default), English (UK), Spanish, French, German, Italian, Portuguese (Brazil), Japanese, Korean or Hindi.

  • voice_promptstring

    How the script is delivered: tone, pacing, accent, emotion. It is not spoken. Default "Say the following."

  • voice_scriptstring

    The exact words to speak when no audio is sent. Send this or audio.

Pricing

Billable unit and optionsRate (USD)
Per second of output video480p$0.02
Per second of output video768p$0.04

Billing notes

  • Billed on the finished video, in whole seconds rounded down.
  • Leaving out resolution gives 768p.

Catalogue rates on 21 Nov 2026. Prices exclude VAT.

View in pricing catalogue