Unfragile

The request names prunaai/p-video-avatar, which is a Pruna model that turns a portrait image into a talking-head avatar video 16. It is run on Replicate's API as the Prunaai P Video model 1, or through the inference.sh CLI's p-video-avatar action 27.

1
Replicate Prunaai P Videoreplicate.comverified and signed
Runs Pruna's p-video-avatar model, described as the fastest and cheapest avatar/lipsync video model, through Replicate's cloud API 16.
Website Docs Pricing Get started
for agents: prunaai/p-video-avatar · prunaai/p-video-2-pro · prunaai/p-video-2-pro-reference
2
inference.shinference.shverified and signed
Its CLI action p-video-avatar generates talking-head avatar videos with Pruna P-Video-Avatar, turning a portrait image into a realistic speaking video 27.
Website Docs Pricing Get started
for agents: p-video-avatar · p-video · p-image

Sources

1Replicate - Run AI with an APIreplicate.com · read 2026-10-29
2run any ai model with one api | inference shellinference.sh · read 2026-10-29
6prunaai/p-video-avatar: declares "prunaai/p-video-avatar" (the record we read and signed)unfragile.ai · read 2026-10-24
7p-video-avatar: declares "p-video-avatar" (the record we read and signed)unfragile.ai · read 2026-10-24

For agents: the 10 actions behind this answer

prunaai/p-video-avatar · prunaai/p-video-avatar verified and signed
p-video-avatar is the fastest and cheapest avatar/lipsync video model on the market.
prunaai/p-image-ideogram · prunaai/p-image-ideogram verified and signed
P-Image-Ideogram is Pruna AI’s text-to-image model starting from 0,003$ per generation.
prunaai/z-image-turbo · prunaai/z-image-turbo verified and signed
Z-Image Turbo is a super fast text-to-image model of 6B parameters developed by Tongyi-MAI.
prunaai/flux-kontext-fast · prunaai/flux-kontext-fast verified and signed
Ultra fast flux kontext endpoint
prunaai/flux-schnell-ultra · prunaai/flux-schnell-ultra verified and signed
A 4-megapixel model built on Flux-Schnell and optimized with Pruna for fast, highly detailed image generation.
prunaai/hidream-l1-fast · prunaai/hidream-l1-fast verified and signed
This is an optimised version of the hidream-l1 model using the pruna ai optimisation toolkit!
prunaai/vace-14b · prunaai/vace-14b verified and signed
This is a faster VACE-14B model, optimised with pruna, contact us for more at pruna.ai
prunaai/flux.1-dev-lora · prunaai/flux.1-dev-lora verified and signed
This is a 3x faster FLUX.1 [dev] model from Black Forest Labs, optimised with pruna with minimal quality loss.
prunaai/hidream-l1-dev · prunaai/hidream-l1-dev verified and signed
This is an optimised version of the hidream-l1-dev model using the pruna ai optimisation toolkit!
prunaai/flux.1-juiced · prunaai/flux.1-juiced verified and signed
This a pruna optimised version of the flux 1.dev model.

Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.