Model · Talking presenters, Video

HunyuanVideo-Avatar

A free-to-download Tencent model that makes a still portrait talk or sing in time with an audio track, with control over emotion and support for more than one character. Needs a strong graphics card; its licence doesn't cover the EU, UK or South Korea.

Go to HunyuanVideo-Avatar’s website

At a glance

Open or closed
Downloadable, with conditions
You can download and run it, but the licence limits use
Weights and code downloadable under Tencent's own community licence, not an open-source licence.
source · checked 12 Oct 2026 · checked on the maker's own page
Licence
Tencent Hunyuan Community License Agreement (Community licence with conditions (e.g. revenue cap))
Products over 100 million monthly active users need a separate licence. Outputs may not be used to improve other AI models; machine-made content must be labelled.
source · checked 12 Oct 2026 · checked on the maker's own page
Can I sell what I make?
Yes, with conditions
  • Not allowed in some countries
  • Only below a revenue limit
  • You must label it as made with the tool
No commercial ban; Tencent claims no rights in outputs. Not licensed in EU/UK/South Korea. revenue_cap stands for the 100M-user cap; brand_label_required approximates 'label AI-generated content'.
source · checked 12 Oct 2026 · checked on the maker's own page
Cost
Free option: Free to download and run yourself
Plans below.
How you use it
On your own computer, In your browser
Single-GPU fp8 and CPU-offload scripts provided. Tencent's Hunyuan Playground hosts a demo. Platform list is our inference (CUDA scripts).
source · checked 12 Oct 2026 · checked on the maker's own page
Skill needed
Technical
source · checked 12 Oct 2026 · checked, but the page was unclear or partial
Made by
Tencent
Hugging Face org is Tencent; no address or country read.
source · checked 12 Oct 2026 · checked on the maker's own page
Status
Available
Inference code and weights released 28 May 2025; single-GPU 10 GB mode added 6 June 2025.
source · checked 12 Oct 2026 · checked on the maker's own page

What you can make with it

audio to videoimage to video

Animates one or more avatar images from audio, with an optional emotion reference image; supports multi-character dialogue.
source · checked 12 Oct 2026 · checked on the maker's own page

Plans and prices

No paid plans listed.

source · checked 12 Oct 2026 · checked on the maker's own page

Rights and safety

Real faces and voices
Copies a real face (animate any portrait photo), no consent step
Runs on any image you supply; no consent step in the software. Licence forbids impersonating someone without consent.
source · checked 12 Oct 2026 · checked, but the page was unclear or partial
Limits
note: Examples use 704px images and 129 frames per sample; frame rate not stated, so clip length isn't given.
source · checked 12 Oct 2026 · checked on the maker's own page

Related

Where you can run it

Hugging Face

Other options for “A talking presenter, a talking version of a photo”

Akool, D-ID, DeeVid AI, DomoAI, Dreamina, Dzine, EchoMimic, Elai.io, fal, FantasyTalking, Hallo2, Hedra, HeyGen, InfiniteTalk, Jogg AI, LivePortrait, Magic Hour, MuseTalk, OmniHuman, SadTalker, VEED, Vidnoz AI, Vozo, WaveSpeedAI

About this entry

Facts last checked 12 Oct 2026 against the maker’s own pages (each fact links to its source). Entry status: draft. Plain-English summary written by Plainly.