TL;DR
Google has released Gemini 3 Flash, delivering frontier-class AI performance at less than a quarter of 3 Pro’s cost. The model scores 90.4% on PhD-level benchmarks while being 3x faster than 2.5 Pro, priced at $0.50 per million input tokens.
Performance Without Compromise
Gemini 3 Flash addresses a persistent developer challenge: choosing between speed and intelligence. The new model delivers frontier-class performance on demanding benchmarks—90.4% on GPQA Diamond (PhD-level science) and 33.7% on Humanity’s Last Exam without tools—rivalling much larger models.
The efficiency gains are substantial. Flash outperforms 2.5 Pro while running 3x faster, according to Artificial Analysis benchmarking. Even at the lowest “thinking level,” 3 Flash often outperforms previous versions with high thinking settings.
Aggressive Pricing for Scale
Google has priced Gemini 3 Flash at $0.50 per million input tokens and $3 per million output tokens—significantly below competitor frontier models. Context caching offers additional 90% cost reductions for repeated token use, while the Batch API provides 50% savings for asynchronous processing.
This pricing opens frontier-class AI to use cases previously cost-prohibitive, potentially democratising access for smaller developers and businesses.
Real-World Applications
Early adopters demonstrate diverse use cases. For coding, 3 Flash matches 3 Pro’s agentic performance (78% on SWE-bench Verified) while operating faster for iterative development. Game developer Astrocade uses it for agentic game creation, generating full plans and executable code from single prompts.
In security, Resemble AI reports 4x faster multimodal analysis for deepfake detection compared to 2.5 Pro, enabling near real-time forensic intelligence. Legal AI company Harvey highlights that the model handles rigorous accuracy demands without sacrificing speed—critical for complex document analysis.
Looking Forward
Gemini 3 Flash is now available through Google AI Studio, Vertex AI, Android Studio, and the new Google Antigravity agentic development platform. For UK businesses building AI applications, the combination of frontier performance, speed, and aggressive pricing removes significant barriers to production deployment.
The model represents Google’s push to make advanced AI accessible at scale—a shift that could accelerate AI adoption across industries where cost previously limited experimentation.