Overview
- Google set pricing at $0.50 per million input tokens and $3.00 per million output tokens, with audio input at $1 per million tokens.
- Company benchmarks report 90.4% on GPQA Diamond, 33.7% on Humanity’s Last Exam without tools, and 81.2% on MMMU Pro, plus a 78% score on SWE-bench Verified.
- The rollout makes Gemini 3 Flash the default in the Gemini app and AI Mode in Search globally, with developer access via Google AI Studio, Antigravity, the Gemini API and CLI, Android Studio, Vertex AI, and Gemini Enterprise.
- Google says the model is roughly three times faster than Gemini 2.5 Pro and uses about 30% fewer tokens on average, with dynamic “thinking” that scales to task complexity.
- Google cites early use by JetBrains, Bridgewater Associates, and Figma and reports processing over one trillion API tokens per day, while analyst Tomasz Tunguz frames the release as major price‑performance deflation versus rivals.