线上展厅
business.online.mosi.cn
apps/showroom
Playground 入口
每个模块保持独立页面和公网接口,点击卡片即可进入对应体验。
语音合成
TTS
文本转语音、音色录制和音色管理。
POST /playground/tts/api/tts/speech
GET/POST /playground/tts/api/tts/voices
/playground/tts/
→
实时识别
ASR
麦克风实时转写,输出说话人和时间戳。
WS /playground/realtime-asr/ws/asr
POST /playground/realtime-asr/api/transcribe_file
/playground/realtime-asr/
→
Voice Agent
VOICE
能听懂、记住并自然回应的实时语音助手,让每一次交流更连贯、更有个性。
POST /playground/voice-agent/v1/sessions
WS /playground/voice-agent/v1/realtime
/playground/voice-agent/
→
字幕生成
SUB
上传音视频,自动切片识别、编辑说话人并导出字幕或烧录视频。
POST /playground/subtitle/api/media
GET /playground/subtitle/api/tasks/{task_id}/export
GET /playground/subtitle/api/tasks/{task_id}/burn
/playground/subtitle/
→
语音质检
QC
通话回放质检、实时告警、规则判定与结果复核。
WS /playground/voice-qc/ws/replay/{call_id}
GET /playground/voice-qc/api/health
/playground/voice-qc/
→
文章配音
VOICE
长文智能拆稿、角色音色匹配,并生成完整 WAV 配音。
POST /playground/article-voice/api/script
POST /playground/article-voice/api/voice-plan
POST /playground/article-voice/api/narrate
/playground/article-voice/
→
实时视频分析
VL
摄像头或本地视频实时抽帧,通过 MOSS-VL 分析画面变化并按需生成判断与应对方案。
POST /playground/frigate-vl-demo/api/direct/analyze
POST /playground/frigate-vl-demo/api/direct/summary
POC4 · /playground/frigate-vl-demo/
→
视频理解与总结
VL
上传长视频,结合语音识别和视觉理解生成分段时间线、关键事件与全片总结。
POST /playground/video-summary/api/jobs
GET /playground/video-summary/api/jobs/{id}/events
POC4 · /playground/video-summary/
→