Google · Gemini 3

Gemini 3 Pro

CurrentVerified

The Pro tier of the original Gemini 3 line: a high-accuracy multimodal model with a 2M-token context window and strong reasoning and coding performance. Still current and widely deployed, sitting just below Gemini 3.5 Pro on quality and price.

profile normalized against the 47-model field

Context window—

Max output—

Output speed—

Affordability—

Capability breadth· 0 of 110%

Capability switches · 0 of 11

Reasoning mode

Tool / function use

Streaming

JSON mode

Structured outputs

Prompt caching

Fine-tuning

Web search

Code execution

Vision input

Audio input

Specifications

Every value carries a primary source and a verification date.

No sourced specifications recorded yet. Specs appear here as they are verified against primary sources.

Benchmarks

Sourced evaluation scores, each verified against its primary source.

GPQA Diamond

GPQA Diamond: 88.9%

88.9 %Verified

SWE-bench Verified

SWE-bench Verified: 74.2%

74.2 %Verified

Humanity's Last Exam (no tools)

Humanity's Last Exam (no tools): 28.4%

28.4 %Verified

AIME 2025

AIME 2025: 91.2%

91.2 %Verified

MMMU-Pro

MMMU-Pro: 78.6%

78.6 %Verified

MMLU-Pro

MMLU-Pro: 87.3%

87.3 %Verified

LiveCodeBench

LiveCodeBench: 76.8%

76.8 %Verified

Loading…