Release notes

This page documents updates to the Gemini API.

September 10, 2025

• Released support for the Embeddings model in Batch API, and added Batch API to the OpenAI compatibility library for even easier ways to get started with batch queries.

September 9, 2025

• Launched Veo 3 and Veo 3 Fast GA, with lower pricing and new options for aspect ratios, resolution, and seeding. Read the Veo documentation for more information.

August 26, 2025

• Launched Gemini 2.5 Image Preview, our latest native image generation model.

August 18, 2025

• Released URL context tool to general availability (GA), a tool for providing URLs as additional context to prompts. Support for using URL context with the gemini-2.0-flash model (available during experimental release) will be discontinued in one week.

August 14, 2025

• Released Imagen 4 Ultra, Standard and Fast models as generally available (GA). To learn more, see the Imagen page.

August 7, 2025

• allow_adult setting in Image to Video generation are now available in restricted regions. See the Veo page for details.

July 31, 2025

• Launched image-to-video generation for the Veo 3 Preview model.
• Released Veo 3 Fast Preview model.
• To learn more about Veo 3, visit the Veo page.

July 22, 2025

• Released gemini-2.5-flash-lite, our fast, low-cost, high-performance Gemini 2.5 model. To learn more, see Gemini 2.5 Flash-Lite.

July 17, 2025

• Launched veo-3.0-generate-preview, the latest update to Veo introducing video with audio generation. To learn more about Veo 3, visit the Veo page.

• Increased rate limits for Imagen 4 Standard and Ultra. Visit the Rate limits page for more details.

July 14, 2025

• Released gemini-embedding-001, the stable version of our text embedding model. To learn more, see embeddings. The gemini-embedding-exp-03-07 model will be deprecated on August 14, 2025.

July 7, 2025

• Launched Gemini API Batch Mode. Batch up requests and send them to process asynchronously. To learn more, see Batch Mode.

June 26, 2025

• The preview models gemini-2.5-pro-preview-05-06 and gemini-2.5-pro-preview-03-25 are now redirecting to the latest stable version gemini-2.5-pro.

• gemini-2.5-pro-exp-03-25 is deprecated.

June 24, 2025

• Released Imagen 4 Ultra and Standard Preview models. To learn more, see the Image generation page.

June 17, 2025

• Released gemini-2.5-pro, the stable version of our most powerful model, now with adaptive thinking. To learn more, see Gemini 2.5 Pro and Thinking. gemini-2.5-pro-preview-05-06 will be redirected to gemini-2.5-pro on June 26, 2025.
• Released gemini-2.5-flash, our first stable 2.5 Flash model. To learn more, see Gemini 2.5 Flash. gemini-2.5-flash-preview-04-17 will be deprecated on July 15, 2025.
• Released gemini-2.5-flash-lite-preview-06-17, a low-cost, high-performance Gemini 2.5 model. To learn more, see Gemini 2.5 Flash-Lite Preview.

June 05, 2025

• Released gemini-2.5-pro-preview-06-05, a new version of our most powerful model, now with adaptive thinking. To learn more, see Gemini 2.5 Pro Preview and Thinking. gemini-2.5-pro-preview-05-06 will be redirected to gemini-2.5-pro on June 26, 2025.

May 20, 2025

API updates:

• Launched support for custom video preprocessing using clipping intervals and configurable frame rate sampling.
• Launched multi-tool use, which supports configuring code execution and Grounding with Google Search on the same generateContent request.
• Launched support for asynchronous function calls in the Live API.
• Launched an experimental URL context tool for providing URLs as additional context to prompts.

Model updates:

• Released gemini-2.5-flash-preview-05-20, a Gemini preview model optimized for price-performance and adaptive thinking. To learn more, see Gemini 2.5 Flash Preview and Thinking.
• Released the gemini-2.5-pro-preview-tts and gemini-2.5-flash-preview-tts models, which are capable of generating speech with one or two speakers.
• Released the lyria-realtime-exp model, which generates music in real time.
• Released gemini-2.5-flash-preview-native-audio-dialog and gemini-2.5-flash-exp-native-audio-thinking-dialog, new Gemini models for the Live API with native audio output capabilities. To learn more, see the Live API guide and Gemini 2.5 Flash Native Audio.
• Released gemma-3n-e4b-it preview, available on AI Studio and through the Gemini API, as part of the Gemma 3n launch.

May 7, 2025

• Released gemini-2.0-flash-preview-image-generation, a preview model for generating and editing images. To learn more, see Image generation and Gemini 2.0 Flash Preview Image Generation.

May 6, 2025

• Released gemini-2.5-pro-preview-05-06, a new version of our most powerful model, with improvements on code and function calling. gemini-2.5-pro-preview-03-25 will automatically point to the new version of the model.

April 17, 2025

• Released gemini-2.5-flash-preview-04-17, a Gemini preview model optimized for price-performance and adaptive thinking. To learn more, see Gemini 2.5 Flash Preview and Thinking.

April 16, 2025

• Launched context caching for Gemini 2.0 Flash.

April 9, 2025

Model updates:

• Released veo-2.0-generate-001, a generally available (GA) text- and image-to-video model, capable of generating detailed and artistically nuanced videos. To learn more, see the Veo docs.

• Released gemini-2.0-flash-live-001, a public preview version of the Live API model with billing enabled.

  • Enhanced Session Management and Reliability

    • Session Resumption: Keep sessions alive across temporary network disruptions. The API now supports server-side session state storage (for up to 24 hours) and provides handles (session_resumption) to reconnect and resume where you left off.
    • Longer Sessions via Context Compression: Enable extended interactions beyond previous time limits. Configure context window compression with a sliding window mechanism to automatically manage context length, preventing abrupt terminations due to context limits.
    • Graceful Disconnect Notification: Receive a GoAway server message indicating when a connection is about to close, allowing for graceful handling before termination.

  • More Control over Interaction Dynamics

  • Configurable Voice Activity Detection (VAD): Choose sensitivity levels or disable automatic VAD entirely and use new client events (activityStart, activityEnd) for manual turn control.

  • Configurable Interruption Handling: Decide whether user input should interrupt the model's response.

  • Configurable Turn Coverage: Choose whether the API processes all audio and video input continuously or only captures it when the end-user is detected speaking.

  • Configurable Media Resolution: Optimize for quality or token usage by selecting the resolution for input media.

  • Richer Output and Features

  • Expanded Voice & Language Options: Choose from two new voices and 30 new languages for audio output. The output language is now configurable within speechConfig.

  • Text Streaming: Receive text responses incrementally as they are generated, enabling faster display to the user.

  • Token Usage Reporting: Gain insights into usage with detailed token counts provided in the usageMetadata field of server messages, broken down by modality and prompt or response phases.

April 4, 2025

• Released gemini-2.5-pro-preview-03-25, a public preview Gemini 2.5 Pro version with billing enabled. You can continue to use gemini-2.5-pro-exp-03-25 on the free tier.

March 25, 2025

• Released gemini-2.5-pro-exp-03-25, a public experimental Gemini model with thinking mode always on by default. To learn more, see Gemini 2.5 Pro Experimental.

March 12, 2025

Model updates:

• Launched an experimental Gemini 2.0 Flash model capable of image generation and editing.
• Released gemma-3-27b-it, available on AI Studio and through the Gemini API, as part of the Gemma 3 launch.

API updates:

• Added support for YouTube URLs as a media source.
• Added support for including an inline video of less than 20MB.

March 11, 2025

SDK updates:

• Released the Google Gen AI SDK for TypeScript and JavaScript to public preview.

March 7, 2025

Model updates:

• Released gemini-embedding-exp-03-07, an experimental Gemini-based embeddings model in public preview.

February 28, 2025

API updates:

• Support for Search as a tool added to gemini-2.0-pro-exp-02-05, an experimental model based on Gemini 2.0 Pro.

February 25, 2025

Model updates:

• Released gemini-2.0-flash-lite, a generally available (GA) version of Gemini 2.0 Flash-Lite, which is optimized for speed, scale, and cost efficiency.

February 19, 2025

AI Studio updates:

• Support for additional regions (Kosovo, Greenland and Faroe Islands).

API updates:

• Support for additional regions (Kosovo, Greenland and Faroe Islands).

February 18, 2025

Model updates:

• Gemini 1.0 Pro is no longer supported. For the list of supported models, see Gemini models.

February 11, 2025

API updates:

• Updates on the OpenAI libraries compatibility.

February 6, 2025

Model updates:

• Released imagen-3.0-generate-002, a generally available (GA) version of Imagen 3 in the Gemini API.

SDK updates:

• Released the Google Gen AI SDK for Java for public preview.

February 5, 2025

Model updates:

• Released gemini-2.0-flash-001, a generally available (GA) version of Gemini 2.0 Flash that supports text-only output.
• Released gemini-2.0-pro-exp-02-05, an experimental public preview version of Gemini 2.0 Pro.
• Released gemini-2.0-flash-lite-preview-02-05, an experimental public preview model optimized for cost efficiency.

API updates:

• Added file input and graph output support to code execution.

SDK updates:

• Released the Google Gen AI SDK for Python to general availability (GA).

January 21, 2025

Model updates:

• Released gemini-2.0-flash-thinking-exp-01-21, the latest preview version of the model behind the Gemini 2.0 Flash Thinking Model.

December 19, 2024

Model updates:

• Released Gemini 2.0 Flash Thinking Mode for public preview. Thinking Mode is a test-time compute model that lets you see the model's thought process while it generates a response, and produces responses with stronger reasoning capabilities.

  Read more about Gemini 2.0 Flash Thinking Mode in our overview page.

December 11, 2024

Model updates:

• Released Gemini 2.0 Flash Experimental for public preview. Gemini 2.0 Flash Experimental's partial list of features includes:
  • Twice as fast as Gemini 1.5 Pro
  • Bidirectional streaming with our Live API
  • Multimodal response generation in the form of text, images, and speech
  • Built-in tool use with multi-turn reasoning to use features like code execution, Search, function calling, and more

Read more about Gemini 2.0 Flash in our overview page.

November 21, 2024

Model updates:

• Released gemini-exp-1121, an even more powerful experimental Gemini API model.

Model updates:

• Updated the gemini-1.5-flash-latest and gemini-1.5-flash model aliases to use gemini-1.5-flash-002.
  • Change to top_k parameter: The gemini-1.5-flash-002 model supports top_k values between 1 and 41 (exclusive). Values greater than 40 will be changed to 40.

November 14, 2024

Model updates:

• Released gemini-exp-1114, a powerful experimental Gemini API model.

November 8, 2024

API updates:

• Added support for Gemini in the OpenAI libraries / REST API.

October 31, 2024

API updates:

• Added support for Grounding with Google Search.

October 3, 2024

Model updates:

• Released gemini-1.5-flash-8b-001, a stable version of our smallest Gemini API model.

September 24, 2024

Model updates:

• Released gemini-1.5-pro-002 and gemini-1.5-flash-002, two new stable versions of Gemini 1.5 Pro and 1.5 Flash, for general availability.
• Updated the gemini-1.5-pro-latest model code to use gemini-1.5-pro-002 and the gemini-1.5-flash-latest model code to use gemini-1.5-flash-002.
• Released gemini-1.5-flash-8b-exp-0924 to replace gemini-1.5-flash-8b-exp-0827.
• Released the civic integrity safety filter for the Gemini API and AI Studio.
• Released support for two new parameters for Gemini 1.5 Pro and 1.5 Flash in Python and NodeJS: frequencyPenalty and presencePenalty.

September 19, 2024

AI Studio updates:

• Added thumb-up and thumb-down buttons to model responses, to enable users to provide feedback on the quality of a response.

API updates:

• Added support for Google Cloud credits, which can now be used towards Gemini API usage.

September 17, 2024

AI Studio updates:

• Added an Open in Colab button that exports a prompt – and the code to run it – to a Colab notebook. The feature doesn't yet support prompting with tools (JSON mode, function calling, or code execution).

September 13, 2024

AI Studio updates:

• Added support for compare mode, which lets you compare responses across models and prompts to find the best fit for your use case.

August 30, 2024

Model updates:

• Gemini 1.5 Flash supports supplying JSON schema through model configuration.

August 27, 2024

Model updates:

• Released the following experimental models:
  • gemini-1.5-pro-exp-0827
  • gemini-1.5-flash-exp-0827
  • gemini-1.5-flash-8b-exp-0827

August 9, 2024

API updates:

• Added support for PDF processing.

August 5, 2024

Model updates:

• Fine-tuning support released for Gemini 1.5 Flash.

August 1, 2024

Model updates:

• Released gemini-1.5-pro-exp-0801, a new experimental version of Gemini 1.5 Pro.

July 12, 2024

Model updates:

• Support for Gemini 1.0 Pro Vision removed from Google AI services and tools.

June 27, 2024

Model updates:

• General availability release for Gemini 1.5 Pro's 2M context window.

API updates:

• Added support for code execution.

June 18, 2024

API updates:

• Added support for context caching.

June 12, 2024

Model updates:

• Gemini 1.0 Pro Vision deprecated.

May 23, 2024

Model updates:

• Gemini 1.5 Pro (gemini-1.5-pro-001) is generally available (GA).
• Gemini 1.5 Flash (gemini-1.5-flash-001) is generally available (GA).

May 14, 2024

API updates:

• Introduced a 2M context window for Gemini 1.5 Pro (waitlist).
• Introduced pay-as-you-go billing for Gemini 1.0 Pro, with Gemini 1.5 Pro and Gemini 1.5 Flash billing coming soon.
• Introduced increased rate limits for the upcoming paid tier of Gemini 1.5 Pro.
• Added built-in video support to the File API.
• Added plain text support to the File API.
• Added support for parallel function calling, which returns more than one call at a time.

May 10, 2024

Model updates:

• Released Gemini 1.5 Flash (gemini-1.5-flash-latest) in preview.

April 9, 2024

Model updates:

• Released Gemini 1.5 Pro (gemini-1.5-pro-latest) in preview.
• Released a new text embedding model, text-embeddings-004, which supports elastic embedding sizes under 768.

API updates:

• Released the File API for temporarily storing media files for use in prompting.
• Added support for prompting with text, image, and audio data, also known as multimodal prompting. To learn more, see Prompting with media.
• Released System instructions in beta.
• Added Function calling mode, which defines the execution behavior for function calling.
• Added support for the response_mime_type configuration option, which lets you request responses in JSON format.

March 19, 2024

Model updates:

• Added support for tuning Gemini 1.0 Pro in Google AI Studio or with the Gemini API.

December 13 2023

Model updates:

• gemini-pro: New text model for a wide variety of tasks. Balances capability and efficiency.
• gemini-pro-vision: New multimodal model for a wide variety of tasks. Balances capability and efficiency.
• embedding-001: New embeddings model.
• aqa: A new specially tuned model that is trained to answer questions using text passages for grounding generated answers.

See Gemini models for more details.

API version updates:

• v1: The stable API channel.
• v1beta: Beta channel. This channel has features that may be under development.

See the API versions topic for more details.

API updates:

• GenerateContent is a single unified endpoint for chat and text.
• Streaming available through the StreamGenerateContent method.
• Multimodal capability: Image is a new supported modality
• New beta features:
  • Function Calling
  • Semantic Retriever
  • Attributed Question Answering (AQA)
• Updated candidate count: Gemini models only return 1 candidate.
• Different Safety Settings and SafetyRating categories. See safety settings for more details.
• Tuning models is not yet supported for Gemini models (Work in progress).