NanoAvatar
New This MonthReal-time talking avatars on older Android phones, without a cloud GPU
- Pricing
- Free
- Platform
- Windows, Android
- Public API
- No
- Open source
- Yes
About NanoAvatar
NanoAvatar generates lifelike talking-avatar video from audio on Android phones, without requiring a cloud GPU for avatar generation. The project includes Android and Python inference code, model weights, and downloadable Android APKs. Experience mode works offline with your own recordings; Conversation mode uses cloud services. The demo page contains recorded videos and download links, not cloud-hosted interactive inference. Compatibility and setup requirements are documented in the repository. Source code: https://github.com/wpydcr/NanoAvatar Model weights: https://huggingface.co/wpydcr/NanoAvatar Recorded demos: https://huggingface.co/spaces/wpydcr/NanoAvatar-Demo Android APKs: https://github.com/wpydcr/NanoAvatar/releases/latest NanoAvatar source code is MIT licensed. Specified weights, default character assets and demo videos use CC BY-NC 4.0; third-party components retain their own licenses. Project author: wpydcr. Contact: [email protected]
Specifications
Everything we hold on this listing.
- Category
- Video
- Pricing
- Free
- Website
- github.com
- Public API
- No
- Open source
- Yes
- Added
- Sep 15, 2026
- Source
- user
How NanoAvatar compares
The closest tools in the directory, matched on shared capability rather than category alone.
| Tool | Pricing | API | Open source |
|---|---|---|---|
| NanoAvatar | Free | No | Yes |
| Creatii AI | Freemium | No | No |
| Magic Hour | Freemium | Yes | Not stated |
| Pika | Paid | Not stated | Not stated |
| Weydub | Freemium | Not stated | Not stated |
Frequently asked
What is NanoAvatar used for?
NanoAvatar generates lifelike talking-avatar video from audio on Android phones, without requiring a cloud GPU for avatar generation. The project includes Android and Python inference code, model weights, and downloadable Android APKs. Experience mode works offline with your own recordings; Conversation mode uses cloud services. The demo page contains recorded videos and download links, not cloud-hosted interactive inference. Compatibility and setup requirements are documented in the repository. Source code: https://github.com/wpydcr/NanoAvatar Model weights: https://huggingface.co/wpydcr/NanoAvatar Recorded demos: https://huggingface.co/spaces/wpydcr/NanoAvatar-Demo Android APKs: https://github.com/wpydcr/NanoAvatar/releases/latest NanoAvatar source code is MIT licensed. Specified weights, default character assets and demo videos use CC BY-NC 4.0; third-party components retain their own licenses. Project author: wpydcr. Contact: [email protected]
Is NanoAvatar free?
NanoAvatar is Free. Completely free, no credit card required.
What platforms does NanoAvatar run on?
NanoAvatar runs on Windows, Android. It has no public API listed, so it's a use-in-the-app tool rather than something to integrate.
Is NanoAvatar open source?
Yes, NanoAvatar publishes its source, so you can self-host it, audit what it does with your data, and avoid per-seat pricing. Browse more open-source options at brofindai.com/tags/open-source.
What are the best alternatives to NanoAvatar?
The closest alternatives are Creatii AI, Magic Hour, Pika, Weydub, all video tools that overlap with NanoAvatar on lip sync. See the full list at brofindai.com/alternatives/nanoavatar-d3bec5.
