Gemini 3 Flash: Google’s New AI Model Outshines Competitors
Google has introduced its latest AI model, Gemini 3 Flash, a lightweight yet powerful addition to its Gemini family. This new model is designed to be faster and more cost‑effective than its predecessors, while still delivering impressive performance in key benchmarks.
What Is Gemini 3 Flash?
Gemini 3 Flash is part of Google’s Gemini 3 lineup, which also includes the Pro and Deep Think models. Although it is a lighter and less expensive variant, it outperforms both Gemini 3 Pro and OpenAI’s GPT‑5.2 in several benchmark tests.
Benchmark Performance
In the “Humanity’s Last Exam” benchmark, Gemini 3 Flash scored 33.7% without tools and 43.5% with search and code execution. For comparison, Gemini 3 Pro achieved 37.5% and 45.8%, while GPT‑5.2 recorded 34.5% and 45.5% respectively.
On the MMMU‑Pro benchmark, which tests multimodal understanding and reasoning, Gemini 3 Flash took the top spot with an 81.2% score, edging out both Gemini 3 Pro and GPT‑5.2.
Cost Efficiency
Gemini 3 Flash costs $0.50 per million input tokens and $3.00 per million output tokens, making it significantly cheaper than Gemini 3 Pro and GPT‑5.2. It also uses 30% fewer tokens on average than Gemini 2.5 Pro, delivering both cost savings and faster processing times.
Availability
For everyday users, Gemini 3 Flash is now the default model in the Gemini app and AI Mode in Google Search. Google claims the model can handle tasks in just a few seconds—such as offering golf‑swing tips from a video or generating a functional app from a series of thoughts.
Developers can access Gemini 3 Flash via the Gemini API in Google AI Studio, the Gemini CLI, and Google Antigravity. Enterprise customers can also leverage it through Vertex AI and Gemini Enterprise.
