Google Launches Gemini 3 Flash, Makes It the Default Model in the Gemini App

December 18, 2025 3 Min Read

Comments Off

Google launches Gemini 3 Flash. Google has officially launched Gemini 3 Flash, a fast, low-cost AI model designed to compete directly with OpenAI’s newest releases. Built on last month’s Gemini 3 technology, the new Flash model is optimized for speed, multimodal understanding, and high-volume workflows, and Google is now making it the default model in the Gemini app and in AI-powered search features worldwide.

The launch comes as Google accelerates its AI rollout amid intensifying competition with OpenAI. Gemini 3 Flash sits between the lightweight Flash 2.5 and the flagship Gemini 3 Pro, delivering significant improvements in reasoning, multimodal tasks, and benchmark scores.

Gemini 3 Flash Delivers Big Performance Gains

According to Google, Gemini 3 Flash dramatically outperforms its predecessor. On Humanity’s Last Exam, a rigorous test of domain expertise:

Gemini 3 Flash: 33.7%
Gemini 3 Pro: 37.5%
Gemini 2.5 Flash: 11%
GPT-5.2: 34.5%

On multimodal reasoning (MMMU-Pro), it even achieved the highest score across competitors with 81.2%.

These improvements make Gemini 3 Flash capable of handling tasks previously reserved for more expensive models.

Consumer Rollout: Gemini 3 Flash Becomes the Default

Google is replacing Gemini 2.5 Flash with Gemini 3 Flash as the default model in the Gemini mobile app, giving users higher-quality outputs at no extra cost. Users can still switch to Gemini 3 Pro for advanced math, deep reasoning, and coding workloads.

The new model is optimized for multimodal use cases:

Upload a video (e.g., pickleball clip) to get technique feedback
Draw a sketch and have the model interpret it
Upload audio for analysis or quiz generation
Request visual answers, including tables and images

Google also added prototype-building tools in the app, letting users design simple app concepts directly with Flash.

Enterprise and Developer Access

Gemini 3 Flash is already being integrated by major enterprise customers, including:

JetBrains
Figma
Cursor
Harvey
Latitude

The model is available through Vertex AI, Gemini Enterprise, and via API for developers. It also works with Antigravity, Google’s new coding tool.

Google highlights that the model is ideal for:

Video analysis
Data extraction
Visual Q&A
High-volume workflows due to speed and token efficiency

Pricing is:

$0.50 per 1M input tokens
$3.00 per 1M output tokens

While slightly higher than Flash 2.5, Google says Gemini 3 Flash is 3× faster and uses 30% fewer tokens for structured reasoning tasks.

A New Phase in the AI Race

The release comes as the AI competition intensifies. After internal concerns about losing consumer traction, OpenAI launched GPT-5.2 and new image models earlier this month. Meanwhile, Google says Gemini adoption continues to climb, with over 1 trillion tokens processed per day across its API ecosystem.

Google executives emphasized that rapid innovation across the industry is pushing all companies to improve.

Receive News Updates and Tutorials Through our Social Media Channels, join: