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

# Google AI Mode API — answers, Markdown and sources

> Ask Google AI Mode a question and get its answer as text and Markdown, with sources, products and citation chips.

`AIMODE` asks Google AI Mode and returns its answer, always with Markdown, and the sources behind it.

| | |
| - | - |
| taskType | `AIMODE` |
| Aliases | `GOOGLE_AIMODE` |
| Reads | `query` (falls back to `prompt`) |
| Credits | 1.25 |
| Country | Optional, default `US`. `CN` asks Google's China market directly, without the US substitution AI Overviews makes. |
| Location | Accepts `payload.location`. See [Regions and location](/engines/capacity.md). |

## Request

```bash theme={"theme":"css-variables"}
curl -X POST https://api.querying.ai/v1/async/task \
  -H "Authorization: Bearer $QUERYING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "taskType": "AIMODE", "payload": { "query": "best CRM for a small team", "country": "US" } }'
```

## Response

Unlike [Google AI Overviews](/engines/google-ai-overviews.md), AI Mode returns a bare answer, not a results-page envelope.

| Field | Type | Description |
| - | - | - |
| `text` | string | The answer. Always present. |
| `markdown` | string | The answer as Markdown. Returned without any flag. |
| `sources` | array | `position`, `url`, `label`, `description`. |
| `shoppingCards`, `inlineProducts` | array | Products in the answer, with the same fields as in AI Overviews. |
| `citationPills` | array | The sources behind each citation chip, with the same fields as in AI Overviews. |
| `videos` | array | Videos the answer shows. |
| `rawContent` | string | The rendered page HTML. Only with `include.rawResponse`. |

Product, chip and video fields appear only when the answer shows them.

For example:

```json theme={"theme":"css-variables"}
{
  "text": "For a small team, start with HubSpot CRM or Pipedrive ...",
  "markdown": "For a small team, start with **HubSpot CRM** or **Pipedrive** ...",
  "sources": [{ "position": 1, "url": "https://www.example.com/crm-guide", "label": "CRM guide for small teams" }]
}
```

When Google loads the page but returns no answer on two fresh sessions, the task fails with `ENGINE_NO_RESULT` instead of waiting.

## Include flags

| Flag | Effect |
| - | - |
| `rawResponse` | Adds `rawContent`. |
| `html` | Adds `html`, the page Google showed. See [HTML pages](/engines/html-pages.md). |
| `analysis` | Adds `analysis`, 1 credit. See [Answer analysis](/engines/answer-analysis.md). |
| `citationAnalysis` | Runs a [Citation Analysis](/research/citation-analysis.md) on the answer. |

`markdown` needs no flag. `searchQueries`, `ads` and `shopping` have no effect. See [Markdown and raw responses](/engines/response-options.md).

## Related

* [Google AI Mode product page](https://querying.ai/en/engines/google-ai-mode)
* [Engine overview](/engines/overview.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.