← All models

Gemini 2.5 Flash Preview Native Audio Dialog

Published Publisher Googlegemini-2.5-flash-preview-native-audio-dialog

Model positioning

A preview model in the Google Gemini 2.5 Flash series, intended for real-time conversation through the Live API and capable of native audio output. It is not the same model as the separately released Thinking variant or text-to-speech model. (aistudio.google.com)

Capabilities and uses

  • Understands audio and generates speech, and can be used to build real-time voice assistants and conversational applications; the Live API also provides audio and video input capabilities. (blog.google)
  • Can adjust speech style, speaking rate, and accent as instructed; supports tool use and function calling, and supports more than 24 languages. (storage.googleapis.com)
  • Can adjust responses based on the user's tone and attempt to ignore background speech or speech not directed at the device, determining when to respond. (storage.googleapis.com)

Limitations and status

  • This is a preview version; its output and interaction behavior are not guaranteed to be consistently accurate and should be tested appropriately when used.
  • A Google announcement states that this model was shut down on October 20, 2025, and its model ID should no longer be relied on to build new services. At the time, the announcement recommended migrating to gemini-2.5-flash-native-audio-preview-09-2025, but that recommendation is historical information; current model availability should be checked separately when selecting a model. (aistudio.google.com)

Pricing

As of November 18, 2026, the historical input or output billing amounts and units for this exact model ID could not be verified on the official pricing page. Prices for other Native Audio versions on the page cannot be directly applied to this model. (aistudio.google.com)

Release and official sources