SenseAvatar

What it does

SenseAvatar turns text scripts, uploaded audio or recordings, uploaded images with motion prompts, and PPT slides into short-form digital-human videos and 24-hour livestream sessions. Short videos include talking-head presentations with synchronized lip movement, expression and gesture, while image-driven videos can run up to 10 minutes. Livestreams can switch products, answer viewer questions, and show close-up shots.

Common uses include producing marketing or course videos without filming presenters, and recreating an existing livestream room from a prior video walkthrough. The 'replicate livestream room' function clones a host's image, voice and products from a recording, and multimodal selection splits footage by sales and interaction metrics. The platform is built on the SenseNova model family.

It suits e-commerce sellers, educators, and enterprise marketing teams that lack filming equipment. It is accessed through the SenseAvatar web platform and client, with a free account that includes a 5-minute trial synthesis credit and free templates. Premium materials and custom avatars or voices require VIP or paid packages, and the library is limited to the vendor's licensed avatars.

What it can do

• Generate talking digital-human videos from text or uploaded audio with lip-synced expressions and motion
• Create image-driven videos up to 10 minutes from an uploaded image, audio, and motion prompt
• Build 24/7 livestream digital hosts from prior livestream footage, with product switching and viewer Q&A
• Replicate an existing livestream room's host image, voice and products from a video walkthrough
• Import PPT slides, generate AI copy, translate, and edit motion segments
• Select over 100 licensed avatars and more than 300 multilingual voices

Example workflows

Product explainer video from a script

Input: a text script and a chosen licensed avatar. Steps: select an avatar and voice, paste the script or upload a PPT, generate the video, and trim motion segments. Output: a lip-synced digital-host video for a product or course.

Recreating a livestream room

Input: a recording of an existing livestream walkthrough. Steps: upload the video, let the replicate function clone host image, voice and products, then schedule and run the stream with automatic product switching. Output: a digital-host livestream that can answer viewer questions.

Before you choose

• Premium templates, custom avatars and voices, and extra synthesis time are paid; prices for these are not published on the page.
• Digital-human content depends on the vendor's avatar and voice library; custom likenesses require a separate custom package.
• Livestream Q&A works best with a configured question set alongside the model's automatic answers.

Listed under

• Creative
• Marketing
• Ecommerce
• Social and video
• Video generation
• Avatar
• Digital human
• Education
• Livestream
• Livestreaming

Get started

New registrations receive 5 minutes of video synthesis time.

1. Create a free account at senseavatar.sensetime.com and log in.
2. Choose a licensed avatar and voice from the library, or upload an image and audio for an image-based digital human.
3. Import a script, PPT, or livestream recording, then generate the video or launch the replicated livestream.

New users receive 5 minutes of video synthesis time at registration, with paid time top-ups, livestream packages, VIP unlocks, and custom avatar packages available in the client. No prices, units or plan tiers are stated, so confirm current plan terms directly with the vendor.

We couldn’t find published prices for SenseAvatar. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• 商汤如影 ↗senseavatar.sensetime.com

Company website

senseavatar.sensetime.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.