Kwaivgi Kling V2 Ai Avatar Pro

Kwaivgi Kling V2 Ai Avatar Pro

Playground

Kling V2 AI Avatar Pro generates high-quality AI avatar videos with clean detail, stable motion, and strong identity consistency—ideal for profiles, intros, and social content. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Kling-v2-ai-avatar-pro — Talking Avatar from Image + Audio

kling-v2-ai-avatar-pro turns a single portrait into a lip-synced talking-head video driven by your own audio. Upload a clear face image, provide a narration or dialogue track, and the model generates a vertical HD avatar clip that speaks and moves naturally on camera.

🌟 Highlights

• Audio-driven performance – Uses your uploaded audio as-is (no TTS), keeping timing, pauses and emotion.
• Photo-real talking avatar – Animates the face, eyes and head while preserving the identity from the reference image.
• One-shot setup – Just an image + audio; no need for video capture or motion recording.
• Portrait-ready output – Produces social-ready vertical video that fits Reels, TikTok, Shorts and story formats.
• Prompt-guided styling (optional) – Use prompt to hint at camera feel or mood (e.g. “soft studio lighting, subtle head movement, gentle smile”).

🔧 Parameters

• audio* – Required. The voice track that drives lip-sync and timing (URL or upload).
• image* – Required. A clear, front-facing portrait of the person to animate.
• prompt – Optional text describing style, expression or camera feel. If omitted, the model uses a neutral talking-head style.

Tip: Use a well-lit, unobstructed face (no heavy motion blur, minimal occlusion) for best identity preservation.

🚀 How to Use

1. Upload audio

   • Clean mono/stereo track, with minimal background noise.
   • Make sure the final edited length matches what you want in the video.

2. Upload image

   • Front or 3/4 view, eyes visible, face not cropped.
   • The avatar’s identity and pose come from this image.

3. (Optional) Add a prompt

   • Guide expression or style, e.g.:

     • “confident presenter in a tech promo, subtle head nods”
     • “friendly customer service tone, warm expression”

4. Run the model

   • The video length is automatically derived from the audio duration.
   • Download the generated talking-head clip and drop it into your editor or directly onto social platforms.

💰 Pricing

Billing is based on audio duration, with a minimum of 5 seconds.

Any clip shorter than 5 seconds is still billed as 5 seconds.

🧠 Tips for Best Results

• Edit your audio first – Remove mistakes, long silences and background noise before upload.
• Match tone to use case – Calm, even delivery for corporate avatars; more expressive reads for ads or UGC.
• Keep framing consistent – Use images with similar head size and framing across a campaign for a unified look.
• Test a few portraits – Small changes in the reference image (lighting, angle) can noticeably change the avatar’s feel.

More Avatar Tools

See our Avatar Tools collection here!

• infinitetalk – WaveSpeedAI’s Infinitetalk generates lip-synced talking-head avatar videos from your scripts or audio, ideal for virtual presenters and explainer content.

• Infinitetalk-muti – WaveSpeedAI’s Infinitetalk-Multi extends the avatar pipeline to multi-speaker / multi-segment scenarios, making it easier to script dialogues, panel shots, or batch avatar content.

• Omni-Human – ByteDance’s Omni-Human 1.5 creates high-fidelity digital humans from images and audio, suitable for realistic virtual hosts, brand ambassadors, and training avatars.

# Submit the task
curl --location --request POST "https://api.wavespeed.ai/api/v3/kwaivgi/kling-v2-ai-avatar-pro" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}" \
--data-raw '{}'

# Get the result
curl --location --request GET "https://api.wavespeed.ai/api/v3/predictions/${requestId}/result" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}"