Skip to main content

Gemini 3.1 Pro

Google

๐Ÿง  Reasoning๐ŸŒ Web Search

Google's flagship. A top-tier reasoner with strong long-context skills at an aggressive price.

Released February 19, 2026

Pricing

Input tokens$2.00/M
Output tokens$12.00/M

Capacity

Context window1.0M tokens

Capabilities

โœ“Reasoning & Planning
โœ“Web Search
โœ—Open Source

Best Scores

GDPval-AA v2962.3 pts
GPQA Diamond94.3%
SWE-bench Verified80.6%
Terminal-Bench 2.173.8%
SWE-bench Pro54.2%

Best For

๐Ÿ”ฌresearch
๐Ÿ“Šanalysis
๐Ÿ’ปcoding

Benchmark Scores

Specialized Skills

GDPval-AA v2

Real paid work from 44 different jobs, such as law, nursing, and software. Judges compare two answers side by side without knowing which model wrote them, and the winner gains rating points. A typical human expert scores 1000, so a higher number means the work was picked over a human more often.

962.3 pts

Provider-reported; no independent run recorded yet.

Knowledge

GPQA Diamond

PhD-level science questions written so you cannot just Google the answer. Tests whether the model can reason about hard science.

94.3%

Independently measured by Artificial Analysis.

Humanity's Last Exam

PhD-level questions across many subjects. Tests deep reasoning on the hardest questions humans can ask.

44.4%

Provider-reported; no independent run recorded yet.

Software Engineering

SWE-bench Verified

Hands the AI bugs from actual software projects and counts how many it fixes. Like a coding job interview, but with real work.

80.6%

Provider-reported; no independent run recorded yet.

Terminal-Bench 2.1

Puts the AI in front of a computer terminal and asks it to finish multi-step tasks on its own. Measures how good an "AI agent" it is.

73.8%

Provider-reported; an independent run by Vals AI (Terminus 2) lands at 70.79%.

SWE-bench Pro

The harder version of the coding test. Bigger codebases, trickier bugs. Scores drop for everyone, so the gaps between models become clearer.

54.2%

Independently measured by Anthropic's comparison table.

Reasoning

Artificial Analysis Intelligence Index

A frequently refreshed overall score made from nine modern tests: real work, tool use, terminal tasks, science, hard questions, and long-context reasoning. It is a scorecard rather than a percent correct.

46.5 pts

Provider-reported; no independent run recorded yet.

Why Choose Gemini 3.1 Pro?

Gemini 3.1 Pro is Google's aggressive flagship. It competes on GPQA reasoning with the best models while underpricing competitors. The 1M context and web search make it excellent for research tasks.

How It Compares

vs Claude Fable 5

Gemini is cheaper and faster; Fable has a slight reasoning edge.

vs GPT-5.6 Sol

Gemini is more aggressive on pricing; Sol leads on terminal/agentic work.

You Might Also Like