Google releases Gemini 3 Flash, a fast and cost-efficient AI model that delivers frontier-level intelligence at Flash-level speed.
- •Achieves 90.4% on GPQA Diamond and 33.7% on Humanity's Last Exam (without tools), rivaling larger frontier models
- •Scores 78% on SWE-bench Verified for coding agent tasks, outperforming both the 2.5 series and Gemini 3 Pro
- •Priced at $0.50/1M input tokens and $3/1M output tokens, running 3x faster than 2.5 Pro
- •Uses 30% fewer tokens on average than 2.5 Pro for everyday tasks while maintaining higher performance
- •Now the default model in the Gemini app and AI Mode in Search; available via Gemini API, Google AI Studio, Vertex AI, and Gemini CLI
This summary was automatically generated by AI based on the original article and may not be fully accurate.
Google Cloud
Databricks
Stripe
Deepmind