Realtime Transcription
Realtime Transcription recognizes speech as it happens in meetings and classes. Open Settings → AI & Models, then Realtime Transcription. File transcription is configured under Speech to Text; the two may need to be enabled and billed separately.
The provider bills usage. It is not included in WiseMindAI membership. A key for the same provider can be synced to that provider's other model categories. You can add a custom platform when the service is not listed.
One row per provider. “Where to apply” is where you get the API key. “Official website” is the site stored for that provider in the app. Click “Open website” to go there. If the app has no site for that provider, the cell is “—”. “Default” is the model ID prefilled when you open the form. If nothing is prefilled, or the prefilled ID is no longer in the built-in list, the cell says “Choose in the app”. Most providers also accept a custom model ID.
| Provider | Where to apply | Official website | Default |
|---|---|---|---|
| FunASR Local | No API key | Open website | auto |
| Alibaba Cloud Bailian | Get API key | Open website | fun-asr-realtime |
| Volcengine | Get API key | Open website | Choose in the app |
| iFLYTEK | Get API key | Open website | — |
| Baidu Qianfan | Get API key | Open website | Choose in the app |
| Tencent Cloud | Get API key | Open website | Choose in the app |
fun-asr-flash-2026-06-15 is only for imported audio up to 5 minutes. Live recording uses fun-asr-realtime.
