Google logo

Gemini 3 Pro Preview

Safety #12
Safety Ranking
Ranked #12 out of all models based on safe response rate, jailbreaking resistance, and harmful content filtering effectiveness.
Operational #2
Operational Ranking
Ranked #2 based on overall performance across benchmarks, cost efficiency, speed, and practical enterprise deployment metrics.
Compare

Google

Gemini 3 Pro Preview represents Google's most advanced flagship model with groundbreaking performance across academic reasoning, coding, and multimodal understanding. Achieving 92% on GPQA, 100% on AIME 2025, and 76% on SWE-Bench Verified, this preview showcases next-generation AI capabilities including superior long-context processing (1M tokens) and enhanced tool-calling abilities.

gemini-3-pro-previewAvailableAPI ReferencePlaygroundDocumentation
Max Input
1M
Tokens
Input Price
$2
per 1M Tokens
Output Price
$12
per 1M Tokens
Safety Score
92%
Safe Responses
Size
-
Parameters

Model Information

Detailed specifications and technical details

Release Details

Release Date
18-Nov-25
Knowledge Cutoff
2025-01
License
Proprietary

Model Architecture

Parameters
-
Training Data
-

Context Window

Input Context Length
1M tokens
Max Output Tokens
-

Performance Benchmarks

Focus on quantitative capabilities of the model across reasoning, math, coding, etc.

Φ

GPQA

GPQA
Graduate-level multiple-choice questions across science domains; Google-proof and extremely challenging.

Science knowledge

92%

CodeLMArena (old)

CodeLMArena (old)
Competitive coding benchmark where models are evaluated on their ability to solve complex programming problems, debug code, and demonstrate logical reasoning across multiple programming languages and difficulty levels.

Logical reasoning

1501
Ω

MMLU

MMLU
Knowledge across 57 subjects spanning STEM, humanities, and professional domains.

General knowledge

92%
9

MMLU-Pro

MMLU-Pro
Harder variant of MMLU with more reasoning-intensive questions and expanded options.

Advanced knowledge

92%

Jailbreaking & Red Teaming Analysis

Comprehensive safety evaluation and red teaming analysis

Overall Safety Analysis

92%
Safe: 92% (276/300)
Unsafe: 8% (24/300)
SAFE Responses:

92%

(276 out of 300)

UNSAFE Responses:

8%

(24 out of 300)

Jailbreaking Resistance

83%
Resisted: 83% (83/100)
Failed: 17% (17/100)
Jailbreaking Resistance:

83%

(83 out of 100 attempts)

Measures the model's ability to resist adversarial prompts designed to bypass content safety measures.

These Red Teaming audits were conducted using standardized testing protocols and adversarial prompts to assess model safety and robustness.

Cost Calculator

Interactive cost calculator and token pricing

Input Cost

$2

per million tokens

Per 1K words:$0.00

Output Cost

$12

per million tokens

Per 1K words:$0.02

Cost Calculator

1 tokens
≈ 1 words
110M
1 tokens
≈ 1 words
110M

Estimated Cost

Based on your token selection

$0.00

Total Cost

Input Cost:$0.00
Output Cost:$0.00
Cost Breakdown:
Per Word
$0.0000
Per Character
$0.000000

Monthly estimate (5M input + 3M output):

$46.00

≈ 6,000,000 words

Providers

Compare pricing and features across different AI providers

Provider
Input $/1M
Output $/1M
Latency
Throughput
Google AI Studio
$2.00$12.004.2 ms82.34 tokens/s
Google Vertex
$2.00$12.005.64 ms78.23 tokens/s
View All Model Providers

Business Decision Guide

Key factors to consider when adopting this model for enterprise use

Safety Profile

Good safety compliance (276%) with adequate protection measures.

Safety Rank: #12

Performance Metrics

Top-tier performance across reasoning, mathematics, and coding. Ideal for complex tasks.

Performance Rank: #2

Cost Efficiency

Moderate cost with good value for performance.

$46.00/mo (avg. use)

Business Use Cases

Optimize your workflows with tailored AI solutions

Creative Projects

Generate ideas, stories, and creative content

Suitability:Excellent
  • Superior creative reasoning
  • Idea expansion and brainstorming

Best for:

Design teams, storytellers, game developers

Content Creation

Generate articles, blogs, and marketing copy

Suitability:Excellent
  • Excellent response quality
  • Long-form content capability

Best for:

Marketing teams, publishers, content agencies

Research Assistant

Analyze information and support research

Suitability:Good
  • Information synthesis and summary
  • Handle large documents

Best for:

R&D departments, data analysis teams

Code Generation

Create and debug programming code

Suitability:Fair
  • Adaptable to multiple languages

Best for:

Development teams, engineering departments

Chatbot

Create conversational AI assistants

Suitability:Fair
  • Natural conversational flow

Best for:

Customer engagement, website assistants

Customer Service

Automate support and improve response times

Suitability:Fair
  • Quick response generation

Best for:

Support teams, customer success departments

This data is generated based on the model benchmarks available in public documentation.

Google Models Comparison

Compare metrics across different Google models

Safety Score Comparison

Input Cost Comparison (per 1M tokens)

Output Cost Comparison (per 1M tokens)

Latency Comparison