models/google/gemini-3-1-flash-lite
G
Google·active

Google: Gemini 3.1 Flash Lite

Google's mid-tier model. Long-context specialist with 1.0M window.

Overall score
4.38
/5.00 · ranked #20
Input
$0.250
per 1M tokens
Output
$1.50
per 1M tokens
Context
1.0M
tokens
Blended
$1.19
3:1 out:in ratio

Price drops, new benchmarks, model updates. Stay current on Google: Gemini 3.1 Flash Lite.

One email per change. Unsubscribe anytime.

modelpicker.aipowered by live benchmark data

Scores by test

Methodology →
Structured Output
5.0
Strategic Analysis
5.0
Constrained Rewriting
4.0
Creative Problem Solving
4.0
Tool Calling
4.0
Faithfulness
4.0
Classification
4.0
Long Context
4.0
Safety Calibration
5.0
Persona Consistency
5.0
Agentic Planning
4.0
Multilingual
5.0
Tabular Data
4.0

What you need to know

Gemini 3.1 Flash Lite is optimized for high-reliability structured data and strategic reasoning at a low price point. Its strongest technical performance lies in structured output, strategic analysis, safety calibration, persona consistency, and multilingual capabilities, all of which score a perfect 5/5. This makes it particularly effective for developers building strict API integrations or enterprise-grade agents that require rigid adherence to formatting and safety guidelines.

The model offers a massive 1.0M token context window, which is rare for a model in this price tier. With a blended cost of $1.19/MTok and input costs as low as $0.250/MTok, it provides significant value for high-volume processing of long documents. While it ranks #22 of 77 overall, its internal scoring suggests it punches above its weight in logic-heavy tasks, despite slight dips in creative problem solving and tool calling.

The pricing is highly competitive for the level of reasoning provided. You are getting top-tier strategic and structural capabilities without the cost overhead of a frontier-class model. The trade-off is a slight decrease in flexibility for nuanced rewriting or complex tool orchestration, though these still maintain a strong 4/5 rating.

Use this model if you need a cost-efficient solution for processing massive datasets, generating reliable JSON/structured outputs, or maintaining strict persona consistency across a large context. Skip this model if your primary requirement is high-level creative synthesis or complex, multi-step tool calling where a 4/5 performance is insufficient.

Strengths — Top 3

Structured Output5.0/5.0
Strategic Analysis5.0/5.0
Safety Calibration5.0/5.0

Relative weaknesses — Bottom 3

Constrained Rewriting4.0/5.0
Creative Problem Solving4.0/5.0
Tool Calling4.0/5.0

Similar models

OGPT-5.4$11.884.54QQwen: Qwen3.5-9B$0.1224.31ZGLM-4.7$1.414.69QQwen: Qwen3.6 27B$2.484.54