Skip to main content
Each voice feature in Enconvo has its own speech model setting, so you can pick a fast streaming model for Dictation and a different one for long files. Choose the model in the feature you are actually using.

​
Choose a model

  1. Open the setting for your feature from the table above.
  2. On a Dictation, Recording, Transcription, or Captioning model page, filter the list with All, Enconvo Cloud, Local, or Custom Providers. In Voice Command and Live Talk, Listen shows the recommended models first, then any other speech model.
  3. Pick a model, then review its language and model options.
Enconvo Cloud models use your Enconvo account, with no API key. Custom Providers use your own API key with that provider. Local models (MLX and Core ML) recognize speech on your Mac once their model is downloaded. The speech provider catalog lists every declared provider, including cloud services and local Parakeet, Qwen, Whisper, SenseVoice, Paraformer, and Nemotron variants. Hardware and availability depend on the runtime and your build.

​
Match the model to the task

​
Test with a short sample

Record a short sample with the language, microphone, and terms you actually use, and review it before you process a long meeting. If it fails, read the error for file size, format, extraction, account, or network details. Local model downloads can be much larger than the app. Check the actual download size and your free space; local speech models don’t all need the same.

Dictionary and Speakers

Teach speech features your words and names.

Permissions

Microphone and system-audio access.

Local Models

Run models on your Mac.

Provider catalog

Every speech provider Enconvo offers.