> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sentivue.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Grok Realtime

> Connect Grok Realtime to SentiVue Talk

Grok Realtime is xAI's Grok Voice Agent, a **speech-to-speech** model that drives **realtime** conversations in SentiVue Talk. It hears the caller and speaks back over one connection, so no separate transcriber or voice provider sits in the path.

This is a different connection from [xAI](/talk/configurations/providers/xai), which supplies Grok voices for a standard voice-and-transcriber pipeline. Connecting one does not connect the other.

## What you need

| Field   | Required |
| ------- | -------- |
| API key | Yes      |

## Getting your API key

1. Sign in at [console.x.ai](https://console.x.ai/).
2. Open **API Keys** for your team.
3. Create a key, name it, and copy the value.

Keys begin with `xai-`.

<Warning>
  xAI requires credits on the account before the API responds. A brand-new team with no credits loaded produces a key that authenticates but fails on every request.
</Warning>

## Adding the connection

1. Go to **Connections** and choose **Grok Realtime**.
2. Paste the API key.
3. Save.

A wrong or expired key is not caught at this point — it first shows up on a live call. Place a test call after saving to confirm the connection works.

## What you can select afterwards

Turn on **Realtime** in [Model Configurations](/talk/configurations/inference-providers) and choose Grok Realtime as the provider.

**Model and voice** — pick from the dropdowns, or type a value if you want one that is not listed. xAI's [models page](https://docs.x.ai/docs/models) lists the current voice-agent models.

## How a speech-to-speech call differs

A speech-to-speech model replaces the transcriber and the voice provider. When a workflow is set to realtime, those two settings no longer apply.

* **The voice comes from the model.** You cannot pair Grok Realtime with a voice from another provider.
* **You still need a language model configured.** It runs the work the realtime model does not — extracting variables and running QA analysis after the call. Without one, those features silently produce nothing.
* **Context compaction does not apply** and is ignored if enabled.

## Troubleshooting

<AccordionGroup>
  <Accordion title="The connection saved fine but calls fail immediately">
    Saving does not check the key against xAI. Confirm the key is active in the console, that you copied it whole including the `xai-` prefix, and that the team it belongs to has credits loaded.
  </Accordion>

  <Accordion title="Requests are rejected even though the key is new">
    xAI needs credits on the account before it serves requests. Load credits on the team that owns the key.
  </Accordion>

  <Accordion title="Variables are not extracted and QA never runs">
    Both need an LLM configured alongside the realtime provider. Check the LLM tab in Model Configurations.
  </Accordion>

  <Accordion title="The workflow still uses a separate voice and transcriber">
    The workflow is not set to realtime. Switch it over on [Model Configurations](/talk/configurations/inference-providers), then pick Grok Realtime — the transcriber and voice settings stop applying at that point.
  </Accordion>
</AccordionGroup>