Choose voice and vision models for an agent

Purpose

Many of Neda's customers send a voice message instead of typing. Pick the models an agent uses to turn speech into text and to create images. These are separate from the agent's main text model. The choices for speaking answers and reading text in documents are saved but do not yet change how the agent works.

Before you start

The custom agent must belong to your current workspace, and you need permission to manage agents. Choose a model with the matching ability. An old selection may remain visible but fail validation when you save.

Steps

1. Open "Agents", find your agent under "My agents" and select it. In the window, open "Agent actions" (the three dots) and choose "Edit agent".
2. Open "Voice". "ASR model" converts uploaded voice messages into text before the agent answers. Choose a model for "ASR model", or leave "Not configured". The TTS and OCR choices are saved, but currently do not change how the agent answers or reads documents.
3. Open "Vision". "Image generation model" creates or edits images from a text request and an optional reference image. Choose a model, or leave "Not configured".
4. If a list says "No compatible models are available yet.", no model with that ability is enabled for you.
5. Select "Save".

Expected result

The agent uses the chosen transcription and image generation models for those jobs. Choosing "Not configured" for a job clears that choice.

Where Data Privacy Mode applies (see Use Data Privacy Mode), only a private model is used for these jobs. With no private transcription model, uploaded voice messages are not transcribed, and images are generated only if a private image model is available.

Troubleshooting