Google DeepMind
Build with Gemini
Try Gemini

Our fastest, most cost-effective 3.5-class model, delivering 350 output tokens per second according to the Artificial Analysis Index

Try Gemini 3.5 Flash-Lite

Handles tasks of varying complexity like coding, UI generation and translation with high quality

A white rounded square icon featuring six small circles arranged in a ring, representing thinking levels, on a soft blue gradient background.A dark rounded square icon featuring six small white circles in a ring representing flexible thinking levels, on a dark blue gradient background.

Flexible thinking levels

Delivers improved reasoning and output quality, allowing users to select the level of thinking they want to use.

A white rounded square icon featuring a black speedometer, representing low latency, on a soft blue gradient background.A dark rounded square icon featuring a white speedometer, representing low latency, on a dark blue gradient background.

Low latency

Tackles tasks that require high throughput.

A white rounded square icon featuring a black code bracket symbol and a sparkle, representing tool use, on a soft blue gradient background.A dark rounded square icon featuring a white code bracket symbol and a sparkle, representing tool use, on a dark blue gradient background.

Tool use

Uses search grounding and instruction following and built-in computer use.

A white rounded square icon featuring a black document with a dollar sign, representing cost-efficiency, on a soft blue gradient background.A dark rounded square icon featuring a white document with a dollar sign, representing cost-efficiency, on a dark blue gradient background.

Cost-efficient

A stronger price-to-performance ratio compared to 3.1 Flash-Lite.

Hands-on

Explore what you can do with Gemini 3.5 Flash-Lite

3.5 Flash-Lite vs. 3.5 Flash latency SxS (API)

3.5 Flash-Lite executes high volume tasks at a lower latency than 3.5 Flash

One prompt, 25 web design options (API)

Working alongside 3.6 Flash as the master agent, 3.5 Flash-Lite instantly generates 25 unique, ready-to-explore web design concepts

Receipt analysis and translation at scale (API)

3.5 Flash-Lite can scale receipt translation and summarization with its multimodal understanding

Rapid game development (API)

3.5 Flash-Lite builds a game by instantly generating and iterating through multiple options

Showcase

“Gemini 3.5 Flash-Lite delivers the intelligence, speed, and cost efficiency needed for Ashler’s agentic retrieval and tool-use tasks. It excels at analyzing fragmented infrastructure records while maintaining context across hundreds of deliverables, supporting our users as they build the next generation of power grids, transportation networks, and data centers.”

Connor Goggins, Co-founder & CTO, Ashler

“Google Gemini 3.5 Flash-Lite is a huge jump from Gemini 3.1 Flash-Lite, punching way above its weight class. Gemini 3.5 Flash-Lite has proven incredibly fast and reliable. Ultimately, the model unlocks direct accessibility, giving users a highly responsive option they can choose on demand for smooth, uninterrupted execution.”

Ashwin Kannan, Principal AI Engineer, Palo Alto Networks

“Gemini 3.5 Flash-Lite landed on the Pareto frontier in our receipt extraction benchmark, offering one of the best tradeoffs we’ve tested between accuracy, latency, and cost. That combination is critical at Ramp, where we need to process receipts quickly and reliably at scale.”

Veeral Patel, Head of Applied AI, Ramp

Performance

3.5 Flash-Lite enables efficient scaling for agentic systems. Across thinking levels, the model significantly outperforms 3.1 Flash-Lite. In fact, on many agentic and coding evals, 3.5 Flash-Lite even outperforms 3 Flash.

BenchmarkNotesGemini 3.5 Flash-LiteGemini 3.1 Flash-LiteGPT-5.4 miniClaude Haiku 4.5
Input price $/1M tokens, no caching$0.30$0.25$0.75$1.00
Output price $/1M tokens$2.50$1.50$4.50$5.00
SWE-Bench Pro (Public) Diverse agentic coding tasks54.2%38.3%54.4%39.5%
Terminal-bench 2.1 Agentic terminal codingTerminus-2 harness54.0%31.0%59.2%44.2%
MLE-Bench Machine Learning Engineering39.2%22.0%——
GDPVal-AA v2 Knowledge workElo11406421171907
OSWorld-Verified Agentic computer use74.0%54.3%72.1%50.7%
CharXiv Reasoning Information synthesis from complex chartsNo tools74.5%73.2%80.3%61.7%
With tools76.5%75.6%——
GDM-MRCR v2 (8-needle) Long context performance128k (average)72.2%60.1%42.7%35.3%
1M (pointwise)21.3%12.3%——

Methodology: deepmind.google/models/evals-methodology/gemini-3-5-flash-lite

Model information

Name
3.5 Flash-Lite
Status
General availability
Input
  • Text
  • Image
  • Video
  • Audio
  • PDF
Output
  • Text
Input tokens
1M
Output tokens
64k
Tool use
  • Function calling
  • Search as a tool
  • Computer use
Best for
  • High-volume, latency-sensitive reasoning tasks
Availability
  • Gemini App
  • Google AI Studio
  • Gemini Enterprise Agent Platform
  • Gemini API
Documentation
View developer docs
Model card
View model card