OpenTalker/SadTalker
Source: https://github.com/opentalker/sadtalker
Structured fields
{
"archived": false,
"checked_at": "2026-10-17T00:00:00.000Z",
"usefulness": "Single portrait image plus audio to talking-head video (CVPR 2023) with head-pose and expression coefficients; widely forked and wrapped, with an A1111 extension. Established baseline for still-image talking heads.",
"source_type": "repository",
"availability": "available",
"dependencies": [
"PyTorch (README install uses CUDA 11.3 builds)",
"ffmpeg"
],
"next_check_at": "2026-11-16T00:00:00.000Z",
"known_limitations": [
"Oldest-dated changelog among the lip-sync entries; README points to an issue for later update tracking, not a maintenance policy."
],
"upstream_updated_at": "2024-06-26T12:51:53.000Z"
}Evidence claims
- [upstream_claim] README: single portrait image + audio = talking head video; CVPR 2023 project; README says the license was updated to Apache 2.0 and the non-commercial restriction removed; latest changelog entry dated 2023.06.12. source
- [sourced_fact] LICENSE file: Apache License 2.0, copyright Tencent, 'except for the third-party components listed below' (list not captured in the fetched text). source
- [upstream_claim] Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| code | Apache-2.0 | permitted | permitted | permitted | unknown | Apache-2.0 except listed third-party components; README disclaimer prohibits using it to harm others, including fraud or portrait-rights infringement. |
| model_weights | unknown | unknown | unknown | unknown | unknown | Pretrained checkpoint license not read; third-party components may differ. |