1 / 3
Swipe to compare

Gemini 3.5 Flash is Google’s latest 3.5-generation lightweight AI model, optimized for autonomous AI agents and high-performance coding tasks thanks to its blazing-fast processing speed and exceptional cost-efficiency. While boasting token generation speeds roughly four times faster than competitor flagship models, it slashes operational costs by half, making large-scale workflows highly economical to process. Equipped with advanced reasoning and multimodal comprehension that surpass previous-generation Pro models, it demonstrates powerful autonomous control. For instance, when integrated with the developer agent platform 'Antigravity 2.0,' it can independently spin up sub-agents and build complex software frameworks in just a few hours. Currently serving as the core engine and default AI model for Google Search, it delivers an innovative web experience by dynamically coding and rendering tailor-made Generative UIs—such as mini-apps and custom web layouts—on the screen in real time to match user requests.

Author
GoogleGoogle
Release Date
2026-05-19
Knowledge Cutoff
2025-01
License
Proprietary
I/O Format
Context Length
10.5M / 66K
API I/O (1M)
$1.5 / $9
How to Use
Output Speed
233 tok/s
Arena Overall
Intelligence Index
55.3
Coding Index
45.0
Math Index
LiveBench
ForecastBench
GPQA Diamond
92.2%
HLE
41.0%
MMLU-Pro
AIME 2025
MATH-500
LB Reasoning
LB Math
LB Data Analysis
LiveCodeBench
LB Coding
LB Agentic
TAU2
95.3%
TerminalBench
40.9%
SciCode
53.1%
IFBench
76.3%
AA-LCR
0.7
Hallucination (HHEM)
Factual Consistency (HHEM)
LB Language
LB Instruction Following