> ## AI Agent Context
> Get the complete documentation index at: https://aistudio.google.com/docs/llms.txt
> Use this file to discover all available pages.
> For the full documentation in a single file, fetch: https://aistudio.google.com/docs/llms-full.txt

# Gemini 3.8 Live Extended Thinking

Gemini 3.8 Live Extended Thinking (`gemini-3.8-live-extended-thinking`) is our
specialized Live API model designed for complex reasoning, multi-step planning,
and long-horizon state tracking during real-time multimodal sessions. It
supports configurable thinking levels (`MINIMAL`, `LOW`, `MEDIUM`, `HIGH`),
thought summaries, and asynchronous function calling.

For sub-second conversational latency where minimal thinking is preferred, see
[Gemini 3.8 Live](/docs/models/gemini-3.8-live).

{% button href="https://aistudio.google.com/live?model=gemini-3.8-live-extended-thinking" variant="ais" %} Try in Google AI Studio {% /button %}

## Documentation  {% id="documentation" %}

Visit the [Live API](/docs/live-api) and
[Thinking in the Live API](/docs/live-api/thinking) guides for full coverage of
reasoning configurations and asynchronous tool integrations.

## gemini-3.8-live-extended-thinking  {% id="gemini-38-live-extended-thinking" %}

{% table variant="model" %}
  * Property
  * Description

  ---

  *     {% section class="model-property" %}
  {% icon name="id_card" /%}Model code

    {% /section %}

  * `gemini-3.8-live-extended-thinking`

  ---

  *     {% section class="model-property" %}
  {% icon name="save" /%}Supported data types

    {% /section %}

  *     **Inputs** Text, images, audio

    **Output** Text and audio


  ---

  *     {% section class="model-property" %}
  {% icon name="token_auto" /%}Token limits{% sup %}[[*]](/docs/tokens){% /sup %}

    {% /section %}

  *     **Input token limit** 131,072

    **Output token limit** 65,536


  ---

  *     {% section class="model-property" %}
  {% icon name="handyman" /%}Capabilities

    {% /section %}

  *     {% section class="three-column-grid" %}
  {% section %}
  **Audio generation**

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  **Batch API**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Caching**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Code execution**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **File search**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Function calling**

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  **Grounding with Google Maps**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Image generation**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Live API**

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  **Search grounding**

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  **Structured outputs**

  {% badge type="not-supported" /%}

  {% /section %}

  {% section %}
  **Thinking**

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  **URL context**

  {% badge type="not-supported" /%}

  {% /section %}

    {% /section %}


  ---

  *     {% section class="model-property" %}
  {% icon name="123" /%}Versions

    {% /section %}

  *     Read the [model version patterns](/docs/models#model-versions) for more details.

    *     Stable: `gemini-3.8-live-extended-thinking`


  ---

  *     {% section class="model-property" %}
  {% icon name="calendar_month" /%}Latest update

    {% /section %}

  * September 2026

  ---

  *     {% section class="model-property" %}
  {% icon name="paid" /%}Pricing

    {% /section %}

  * [Pricing](/docs/pricing#gemini-3.8-live)
{% /table %}

## Key considerations and behavior  {% id="key-considerations" %}

When building with `gemini-3.8-live-extended-thinking`, consider the following
architectural requirements:

*   **Default thinking level**: Defaults to `thinkingLevel="HIGH"` to maximize
  reasoning depth and instruction adherence. You can adjust this to `MEDIUM`,
  `LOW`, or `MINIMAL` in `ThinkingConfig` to balance latency and reasoning
  depth. See
  [Configure thinking levels](/docs/thinking#thinking-levels).
*   **Asynchronous function calling required**: Because reasoning takes
  additional time before and after tool execution,
  `gemini-3.8-live-extended-thinking` enforces asynchronous (`NON_BLOCKING`)
  function calling. All custom function declarations must include `"behavior":
"NON_BLOCKING"`. This ensures the model can continue conversing with the
  user while tools execute in the background. See
  [Asynchronous function calling](/docs/live-api/tools#async-function-calling).
*   **Thought summaries**: You can set `include_thoughts=True` in
  `ThinkingConfig` to receive streaming text summaries of the model's internal
  reasoning (`part.thought == True`) before audio playback begins. See
  [Receive thought summaries](/docs/thinking#summaries).
*   **Proactive audio and affective dialogue**: Both features are fully
  supported, allowing the model to reason about emotional tone and determine
  when to speak or remain silent. See
  [Proactive audio](/docs/live-api/capabilities#proactive-audio) and
  [Affective dialogue](/docs/live-api/capabilities#affective-dialog).

For a detailed feature comparison across all Live API models, see the
[Model comparison](/docs/live-api/capabilities#model-comparison) table in the
capabilities guide.