> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.jambonz.org/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.jambonz.org/_mcp/server.

# Google (AI Studio)

Google AI Studio is the simplest way to use Gemini — sign in with a Google account, get an API key, done. The free tier has generous monthly limits, making this the easiest entry point for prototyping.

If you need enterprise data controls, GCP IAM, or Vertex regional residency, use [Vertex AI — Gemini](/guides/features/bring-your-own-llm/vertex-gemini) instead. Same models, different infrastructure path.

## Get credentials

1. Sign in at [https://aistudio.google.com](https://aistudio.google.com) with your Google account.
2. Click **Get API key** in the sidebar (direct link: [https://aistudio.google.com/apikey](https://aistudio.google.com/apikey)).
3. Click **Create API key**, optionally tie it to a Google Cloud project, copy the key.

> **Tip**
>
> AI Studio API keys do not expire by default but can be revoked from the same page. Each key is tied to a Google Cloud project for billing purposes.

## Configure in jambonz

In the portal: **Account → LLM Services → + Add LLM Service → Google Gemini**.

**`API Key`** `string` — required

The key from AI Studio.

---

Click **Test** to verify.

## Use in an agent verb

```js
session.agent({
  llm: {
    vendor: 'google',
    model: 'gemini-2.5-flash',
    llmOptions: {
      systemPrompt: 'You are a helpful voice assistant.',
    },
  },
  stt: { vendor: 'deepgram', language: 'en-US' },
  tts: { vendor: 'cartesia', voice: 'sonic-english' },
  turnDetection: 'krisp',
  bargeIn: { enable: true },
  actionHook: '/agent-complete',
}).send();
```

## Available Models

See Google's [Gemini model catalog](https://ai.google.dev/gemini-api/docs/models) for the full list and current pricing. Common picks for voice agents:

| Model                   | When to use                                                     |
| ----------------------- | --------------------------------------------------------------- |
| `gemini-2.5-flash`      | Default for voice — fast, cheap, tool-capable, 1M-token context |
| `gemini-2.5-pro`        | Higher reasoning quality                                        |
| `gemini-2.0-flash-lite` | Cheapest option; no vision support                              |

## Quirks & errors

> **Note**
>
> **`google` is the AI Studio path, not Vertex.** If your organization requires GCP IAM, regional data residency, or pre-paid commitments, you want [`vertex-gemini`](/guides/features/bring-your-own-llm/vertex-gemini) instead. The model catalog overlaps but the auth and infra differ.

> **Note**
>
> **Free-tier rate limits.** AI Studio's free tier caps requests-per-minute pretty aggressively. For production traffic, upgrade to a paid tier in your Google Cloud billing account or switch to Vertex AI for higher quotas.

> **Warning**
>
> **`400 API_KEY_INVALID`** — the key has typos or was revoked. Regenerate at [aistudio.google.com/apikey](https://aistudio.google.com/apikey).

> **Warning**
>
> **`429 RESOURCE_EXHAUSTED`** — you've hit a per-minute or per-day quota. Switch to a paid tier or distribute traffic across multiple keys.