MiGPT
Add language-model conversations to XiaoAI speakers through a server-side bridge.
Platform / multi-device project · Family mi-gpt · Reviewed 2026-10-20
What you need
- Method
- Service bridge
- Difficulty (editorial judgment)
- Intermediate
- Disassembly
- Not required
- Soldering
- Not required
- Cloud API / service
- Xiaomi service required
- Cost amount
- Not priced / UNKNOWN
Possible additional costs
- Always-on host
- Model/TTS API
Prices have not been checked. An open-source license does not make hardware, hosting, electricity or cloud services free.
Risks and recovery
Data-loss risk: Possible; Bricking risk: Workflow-dependent. Route-based guidance, not hardware-tested; these outcomes are not inevitable for every device.
Stop the Docker/Node service to stop bridging; the original speaker firmware is retained.
Recovery documentation (no verified link)
Evidence and maintenance
E1 · Source code & deployment docs
The README contains installation, model connection and device compatibility instructions and explicitly announces the end of maintenance.
Catalog hands-on testing: not performed. Independent reproduction: not verified.
- Maintenance
- Archived
- Last GitHub push
- 2026-04-04T07:53:49Z
- Repository license
- MIT
An unarchived repository or recent push does not prove a model works. Mirror migration does not mean development has stopped.
A root license or README declaration does not establish the same license for firmware, models, brands and third-party components.
Sources
- Upstream documentation · 2026-10-20 · Content reviewed
Original sources may be in Chinese or English. Review means reading public material; recheck upstream models, services and firmware before acting.