> ## AI Agent Context
> - Full documentation index: https://aistudio.google.com/llms.txt
> - Gemini Docs MCP server (consistent live info): https://gemini-api-docs-mcp.dev
> - Coding agent skills & setup guide: https://aistudio.google.com/docs/coding-agents.md

# Gemini 3.8 Live

Gemini 3.8 Live (`gemini-3.8-live`) is the default option for low-latency,
audio-to-audio voice agent experiences and real-time dialogue without
reasoning-induced delays. It supports interleaved reasoning, asynchronous and
synchronous function calling, full-session client content updates, and built-in
proactive audio.

For complex agentic tasks requiring multi-step reasoning and configurable
thinking budgets, see
[Gemini 3.8 Live Extended Thinking](/docs/models/gemini-3.8-live-extended-thinking).

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

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

Visit the [Live API](/docs/live-api) guide for full coverage of features,
streaming configurations, and tool integrations.

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

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

  ---

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

    {% /section %}

  * `gemini-3.8-live`

  ---

  *     {% 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**](/docs/speech-generation)

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  [**Code execution**](/docs/code-execution)

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

  {% /section %}

  {% section %}
  [**Context caching**](/docs/caching)

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

  {% /section %}

  {% section %}
  [**File search**](/docs/file-search)

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

  {% /section %}

  {% section %}
  [**Function calling**](/docs/function-calling)

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  [**Grounding with Google Maps**](/docs/maps-grounding)

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

  {% /section %}

  {% section %}
  [**Grounding with Google Search**](/docs/google-search)

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  [**Image generation**](/docs/image-generation)

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

  {% /section %}

  {% section %}
  [**Live API**](/docs/live-api)

  {% badge type="supported" /%}

  {% /section %}

  {% section %}
  [**Structured outputs**](/docs/structured-output)

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

  {% /section %}

  {% section %}
  [**Thinking**](/docs/thinking)

  {% badge label=Supported (interleaved reasoning) type="supported" /%}

  {% /section %}

  {% section %}
  [**URL context**](/docs/url-context)

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

  {% /section %}

    {% /section %}


  ---

  *     {% section class="model-property" %}
  {% icon name="speed" /%}Consumption options

    {% /section %}

  *     {% section class="three-column-grid" %}
  {% section %}
  [**Batch API**](/docs/batch-api)

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

  {% /section %}

  {% section %}
  [**Flex inference**](/docs/flex-inference)

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

  {% /section %}

  {% section %}
  [**Priority inference**](/docs/priority-inference)

  {% 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`


  ---

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

    {% /section %}

  * September 2026

  ---

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

    {% /section %}

  * [Model card](https://deepmind.google/models/model-cards/gemini-3-8-audio/)

  ---

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

    {% /section %}

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

## Migrating from Gemini 3.1 Flash Live  {% id="migrating" %}

Gemini 3.8 Live delivers ultra-low latency audio-to-audio interactions and
expands support for asynchronous workflows. When migrating from
`gemini-3.1-flash-live-preview`, review the following updates:

*   **Model string**: Update your model string from
    `gemini-3.1-flash-live-preview` to `gemini-3.8-live`.
*   **Thinking configuration**: `thinking_level` is not supported in
    `gemini-3.8-live` because interleaved reasoning is built in. Omit
    `thinking_level` and `thinking_config` from your session setup. If your
    application requires configurable background reasoning (`low`, `medium`, or
    `high`), use
    [`gemini-3.8-live-extended-thinking`](/docs/models/gemini-3.8-live-extended-thinking).
    See [Thinking in the Live API](/docs/live-api/thinking).
*   **Asynchronous function calling**: Asynchronous execution (`behavior:
    NON_BLOCKING`) is the default function calling mode. You can still use
    synchronous blocking mode for backward compatibility by setting `behavior:
    BLOCKING` on your tool declarations. Function scheduling options (`SILENT`,
    `WHEN_IDLE`, and `INTERRUPTED`) are also supported. See
    [Asynchronous function calling](/docs/live-api/tools#async-function-calling).
*   **Client content updates**: `send_client_content` is supported throughout
    the entire session lifecycle (both initial history seeding and mid-turn
    context updates) with explicit roles (`user` or `model`). Setting
    `turn_complete=true` unconditionally interrupts active model generation. If
    you send content without `turn_complete`, the server waits for subsequent
    messages before responding. See
    [Incremental content updates](/docs/live-api/capabilities#incremental-updates).
*   **Proactive audio**: Proactive audio is permanently enabled. Setting
    `proactive_audio: false` returns an error, so omit `proactive_audio` if you
    are migrating code that disables it.
*   **Affective dialogue**: Affective dialogue is removed from the API. Remove
    any `enable_affective_dialog` configurations from your setup.
*   **Turn coverage**: Defaults to
    [`TURN_INCLUDES_AUDIO_ACTIVITY_AND_ALL_VIDEO`](/api-reference/live#RealtimeInputConfig.TurnCoverage).
    Video frames are sent to the model by default, so send frames only when
    needed to manage context and cost.
*   **Response modalities**: Audio is the only supported response modality
    (`response_modalities=["AUDIO"]`). Enable `output_audio_transcription` if
    your application requires a text transcript. See
    [Audio transcriptions](/docs/live-api/capabilities#audio-transcription).

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

