P Video Avatar

P Video Avatar turns a single portrait image into a realistic talking-head video in seconds. Provide a front-facing photo and either a text script (spoken by one of 30 voices across 10 languages) or an audio file to lip-sync. The model prioritizes speed and affordability while maintaining natural mouth shapes, facial micro‑expressions, and clear speech. How it works - Input a portrait image (jpg, jpeg, png, webp) - Add either a voice_script (text-to-speech) or an audio file (lip‑sync). If both are provided, audio takes precedence. - Optionally guide performance with voice_prompt (tone, pacing, emotion) and describe on‑screen behavior with video_prompt. - Choose resolution (720p for best value, 1080p for extra detail). Set a seed for reproducibility. - Receive an MP4 video URL with embedded speech. Key inputs (high level) - image: Required portrait image to animate - audio: Optional uploaded audio to drive lip‑sync - voice, voice_script, voice_language: TTS settings for scripted speech - voice_prompt: Style, tone, pacing, emotion (not spoken aloud) - video_prompt: Visual behavior description (e.g., “the person is talking”) - resolution: 720p or 1080p - seed: Random seed for repeatable results - disable_prompt_upsampling: Use your exact video_prompt without enhancement - disable_safety_filter: Skips unsafe-content checks (use with care) Voices and languages - 30 curated male/female voices - 10 languages, including English (US/UK), Spanish, French, German, Italian, Portuguese (Brazil), Japanese, Korean, and Hindi Output - MP4 video at 720p or 1080p with the speech baked into the audio track Best practices for quality - Use a clear, front‑facing portrait with good lighting and minimal occlusion - Provide clean, well‑paced audio for tight lip‑sync - Keep prompts specific: use voice_prompt for delivery style, video_prompt for on‑screen behavior - Choose 720p for rapid, low‑cost iterations; switch to 1080p for final polish Pricing (indicative) - Billed per second of output: 720p ≈ $0.025/s, 1080p ≈ $0.045/s - Example: a 10‑second clip at 720p ≈ $0.25 What you can build - Educational explainers and training intros with a consistent on‑screen presenter - Product, marketing, and social videos from a brand photo or mascot - Multilingual/localized announcements from a single source portrait - Podcast/audiobook visuals that lip‑sync to existing recordings Notes and limitations - Optimized for head‑and‑shoulders framing; large gestures or full‑body motion are limited - Strong identity preservation requires a clean, front‑facing image - Use rights‑cleared images and audio and respect consent and local publicity laws