admin_api-en.md

Admin API Reference

Base URL: http://<host>:<listenport>

---

1. Get global configuration

GET /api/admin/config

Returns the global configuration used when the server started, including CLI options.

Request parameters: None.

Response:

{
  "code": 0,
  "msg": "ok",
  "data": {
    "config": {
      "fps": 25,
      "l": 10,
      "m": 8,
      "r": 10,
      "model": "wav2lip",
      "avatar_id": "wav2lip256_avatar1",
      "data_path": "data/avatars",
      "batch_size": 16,
      "modelres": 192,
      "modelfile": "",
      "customvideo_config": "",
      "tts": "edgetts",
      "REF_FILE": "zh-CN-YunxiaNeural",
      "REF_TEXT": null,
      "TTS_SERVER": "http://127.0.0.1:9880",
      "transport": "webrtc",
      "push_url": "http://localhost:1985/rtc/v1/whip/?app=live&stream=livestream",
      "max_session": 1,
      "listenport": 8010,
      "customopt": []
    }
  }
}

Configuration fields

Field	Type	Default	Description
fps	int	25	Video frame rate
l	int	10	Audio buffer parameter
m	int	8	Audio buffer parameter
r	int	10	Audio buffer parameter
model	string	"wav2lip"	Avatar model: musetalk / wav2lip / ultralight
avatar_id	string	"wav2lip256_avatar1"	Default avatar ID
data_path	string	"data/avatars"	Avatar data directory
batch_size	int	16	Inference batch size
modelres	int	192	Model resolution
modelfile	string	""	Custom model file path
customvideo_config	string	""	Custom action configuration JSON path
tts	string	"edgetts"	TTS plugin
REF_FILE	string	"zh-CN-YunxiaNeural"	TTS reference file or voice model ID
REF_TEXT	string	null	TTS reference text
TTS_SERVER	string	"http://127.0.0.1:9880"	TTS server address
transport	string	"webrtc"	Output transport: rtcpush / webrtc / rtmp / virtualcam
push_url	string	—	RTCPush destination URL
max_session	int	1	Maximum number of sessions
listenport	int	8010	HTTP listening port
customopt	array	[]	Parsed custom action configuration

---

2. List active sessions

GET /api/admin/sessions

Returns all active sessions with their status and configuration.

Request parameters: None.

Response:

{
  "code": 0,
  "msg": "ok",
  "data": {
    "sessions": [
      {
        "sessionid": "uuid-string",
        "speaking": true,
        "recording": false,
        "model": "musetalk",
        "avatar_id": "avatar1",
        "REF_FILE": "zh-CN-YunxiaNeural",
        "transport": "webrtc",
        "batch_size": 16,
        "customopt": []
      }
    ]
  }
}

Session fields

Field	Type	Description
sessionid	string	Unique session ID
speaking	bool	Whether the session is currently speaking
recording	bool	Whether the session is currently recording
model	string	Avatar model used by the session
avatar_id	string	Avatar ID used by the session
REF_FILE	string	TTS reference file or model ID
transport	string	Output transport
batch_size	int	Inference batch size
customopt	array	Custom action configuration