modelgrep
G

Google: Gemini 3 Flash Preview

google/gemini-3-flash-preview

22nd smartest of 178ReasoningToolsJSONVisionAudio
Use via OpenRouter ↗
Intelligence
46.4
22nd of 178
Design Elo
1265
3D
Speed
66
122nd fastest
Latency
1.2s
first token
Input price
$0.500
162nd cheapest
Context
1.0M
66K max out

How it compares

Smarter than88%
of all ranked models
Faster than59%
of all ranked models
Cheaper than46%
of all ranked models

Overview

Gemini 3 Flash Preview is a high speed, high value thinking model designed for agentic workflows, multi turn chat, and coding assistance. It delivers near Pro level reasoning and tool...

Benchmarks

independent · via OpenRouter
Artificial Analysis85th percentile
Intelligence Index
46.4
Coding Index
42.6
Agentic Index
49.7
GPQA Diamond
90%
Humanity's Last Exam
35%
SciCode
51%
Tau²-Bench (agentic)
80%
Design Arena · Elo9,441 tournaments
3D
1265
Website
1241
codecategories
1241
Game Dev
1235
godotgamedev
1220
webapps
1200
mobileapps
1196
fullstack
1141
agenticslides(python-pptx)
1075
agenticslides
1073
androidnative
1068

Providers & pricing (2)

ProviderIn $/MOut $/MUptime
Google AI Studiocache$0.500$3.0099.6%
Googlecache$0.500$3.0098.7%

Specifications

Context window1.0M
Max output66K
Knowledge cutoff
Input modalitiestext, image, file, audio, video
Output modalitiestext
Prompt cachingSupported
Cache read price$0.050/M
ModeratedNo

Gemini 3 Flash Preview FAQ

How much does Gemini 3 Flash Preview cost?

Gemini 3 Flash Preview costs $0.500 per million input tokens and $3.00 per million output tokens via OpenRouter, making it 162nd cheapest of 298 paid models.

How smart is Gemini 3 Flash Preview?

Gemini 3 Flash Preview scores 46.4 on the Artificial Analysis Intelligence Index, ranking 22nd of 178 benchmarked models, with a GPQA Diamond score of 90%.

How fast is Gemini 3 Flash Preview?

Gemini 3 Flash Preview generates around 66 tokens per second with 1.2s time-to-first-token (p50), the 122nd fastest tracked model.

What is Gemini 3 Flash Preview's context window?

Gemini 3 Flash Preview supports a 1.0M-token context window and can output up to 66K tokens. It accepts text, image, file, audio, video input.

Compare head-to-head