Gemini 3 Pro

Name: Gemini 3 Pro
Author: Google

Multimodal

Google

Gemini 3 Pro is the first model in the new Gemini 3 series. Best suited for complex tasks requiring broad world knowledge and advanced reasoning across modalities. Gemini 3 Pro uses dynamic thinking by default to reason over prompts and has a 1 million token input context window with up to 64k output tokens.

Key Specifications

Parameters

Context

1.0M

Release Date

November 17, 2025

Average Score

70.5%

Results Blog

Timeline

Key dates in the model's history

Announcement / Last Update

November 17, 2025

Today

October 5, 2026

Technical Specifications

Parameters

Training Tokens

Knowledge Cutoff

January 1, 2025

Family

Capabilities

MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)

$2.00

Output (per 1M tokens)

$12.00

Max Input Tokens

1.0M

Max Output Tokens

65.5K

Supported Features

Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Programming

Programming skills tests

SWE-Bench Verified

Single attempt • Self-reported

76.2%

Reasoning

Logical reasoning and analysis

GPQA

Without tools • Self-reported

92.0%

Other Tests

Specialized benchmarks

AIME 2025

With code • Self-reported

100.0%

MMMLU

• Self-reported

92.0%

ARC-AGI v2

ARC Prize Verified • Self-reported

31.1%

CharXiv-R

Self-reported by the model provider • Self-reported

81.4%

FACTS Grounding

Self-reported by the model provider • Self-reported

70.5%

Global PIQA

Self-reported by the model provider • Self-reported

93.4%

Humanity's Last Exam

With search and code execution • Self-reported

45.8%

LiveBench

2026-01-08, High • Verified

73.4%

LiveCodeBench Pro

Elo Rating • Self-reported

81.3%

MathArena Apex

Self-reported by the model provider • Self-reported

23.4%

MMMU-Pro

Self-reported by the model provider • Self-reported

81.0%

MRCR v2 (8-needle)

1M (pointwise) • Self-reported

26.3%

ScreenSpot Pro

Self-reported by the model provider • Self-reported

72.7%

SimpleQA

Self-reported by the model provider • Self-reported

72.1%

t2-bench

Self-reported by the model provider • Self-reported

85.4%

Terminal-Bench 2.0

Terminus-2 agent • Self-reported

54.2%

VideoMMMU

Self-reported by the model provider • Self-reported

87.6%

License & Metadata

License

proprietary

Announcement Date

November 17, 2025

Last Updated

November 17, 2025

Articles about Gemini 3 Pro

Google Handles a Billion Health Questions a Day. Should We Be Worried?

At The Check Up 2026, Google revealed the scale of its health AI ambitions — from Fitbit medical records to breast cancer screening. But controversies linger.

March 23, 2026

8 min

Similar Models

All Models

Gemini 2.5 Pro Preview 06-05

Google

Best score:0.9 (GPQA)

Released:Jun 2025

Price:$1.25/1M tokens

Gemini 3.1 Pro

Google

Best score:0.9 (GPQA)

Released:Feb 2026

Price:$2.50/1M tokens

Gemini 3 Flash

Google

Best score:0.9 (GPQA)

Released:Dec 2025

Price:$0.50/1M tokens

Gemini 3.1 Flash-Lite

Google

Best score:0.9 (GPQA)

Released:Mar 2026

Price:$0.25/1M tokens

Gemini 2.5 Pro

Google

Best score:0.8 (GPQA)

Released:May 2025

Price:$1.25/1M tokens

Gemini 3.7 Flash

Google

Released:Aug 2026

Price:$0.75/1M tokens

Gemini 3.8 Flash

Google

Released:Sep 2026

Price:$0.75/1M tokens

Gemini 2.0 Flash-Lite

Google

Best score:0.5 (GPQA)

Released:Feb 2025

Price:$0.07/1M tokens

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.