Gemini 3 Flash
Preview
This product or feature is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of theService Specific Terms, and theAdditional Terms for Generative AI Preview Products. You can process personal data for this product or feature as outlined in theCloud Data Processing Addendum, subject to the obligations and restrictions described in the agreement under which you access Google Cloud. Pre-GA products and features are available "as is" and might have limited support. For more information, see thelaunch stage descriptions.
Gemini 3 Flash combines Gemini 3 Pro's reasoning capabilitieswith the Flash line's levels on latency, efficiency, and cost. It not onlyenables everyday tasks with improved reasoning, but is designed to tackle themost complex agentic workflows.
Gemini 3 Flash uses several new features to improve performance,control, and multimodal fidelity:
Thinking level: Use the
Note: If you used a thinking budget ofthinking_levelparameter to control the amountof internal reasoning the model performs (minimal,low,medium, orhigh) to balance response quality, reasoning complexity, latency, andcost. Thethinking_levelparameter replacesthinking_budgetforGemini 3 models.0with Gemini 2.5 Flash,set your thinking level toMINIMALfor similar latency and cost; however,you still need to handle thought signatures when using theminimalthinking level.For details on the different thinking levels, seeThinking.
Thought signatures: Stricter validation ofthought signaturesimproves reliability in multi-turn function calling.
Media resolution: Use the
media_resolutionparameter (low,medium,high, orultra high) to control vision processing for multimodal inputs,impacting token usage and latency. SeeGet started withGemini 3for default resolution settings.- Theultra high media resolution level is only available for the
IMAGEmodality. - PDF token counts will be listed under the
IMAGEmodality instead oftheDOCUMENTmodality inusage_metadata.
- Theultra high media resolution level is only available for the
Multimodal function responses: Function responses can now includemultimodal objects like images and PDFs in addition totext.
Streaming Function calling:Stream partial function call argumentsto improve user experience during tool use.
For more information on using these features, seeGet started withGemini3.
Try inVertex AIView inModel Garden(Preview) Deploy example app
| Model ID | gemini-3-flash-preview | |
|---|---|---|
| Supported inputs & outputs |
| |
| Token limits |
| |
| Capabilities | ||
| Consumption options |
| |
| SeeConsumption options for more information. | ||
| Technical specifications | ||
| Images |
| |
| Documents |
| |
| Video |
| |
| Audio |
| |
| Parameter defaults |
| |
| Supported regions | ||
Model availability |
| |
| SeeDeployments and endpoints for more information. | ||
| Knowledge cutoff date | January 2025 | |
| Versions |
| |
| Supported languages | SeeSupported languages. | |
| Pricing | SeePricing. | |
Except as otherwise noted, the content of this page is licensed under theCreative Commons Attribution 4.0 License, and code samples are licensed under theApache 2.0 License. For details, see theGoogle Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2026-02-19 UTC.