Gemini 3.1 Pro Overview

Gemini 3.1 Pro: Google's Pro Model

Gemini 3.1 Pro was Google's Pro model on Magicdoor, with a 1 million token context window. It is no longer in the current chat model picker. This page preserves its historical capabilities and pricing.

Pricing

• Prompt tokens: $2.00 per million
• Completion tokens: $12.00 per million

These were Gemini 3.1 Pro's listed rates while it was available. For current model choices and prices, see the model lineup.

What It's Good For

Its massive context window was the main draw for tasks such as:

• Analyze an entire codebase at once
• Process multiple long documents together
• Maintain context across a very long conversation

Gemini 3.1 Pro could handle those large inputs where smaller-context models needed trimming.

It supported vision, web search, reasoning mode, and PDF analysis. Its main gaps compared with other flagships were no Canvas support and no code interpreter.

Where Gemini 3.1 Pro Fit

Gemini 3.1 Pro suited work with lots of text:

• Research projects requiring multiple source documents
• Code review across large projects
• Long-form analysis where context matters

When to Use Something Else

• For code execution: GPT-6.1 Sol has code interpreter
• For Canvas editing: GPT-6.1 Sol or Claude Sonnet 5.5 support it
• For polished writing: Claude Opus 5.5 excels at polished prose
• For a low-cost Anthropic model: Claude Haiku 5.5 is available in the chat lineup

Getting Started

Gemini 3.1 Pro is no longer selectable from the model dropdown. For a current Anthropic chat model, choose Claude Haiku 5.5.

For current large-context work, check the model lineup before choosing a model; Gemini 3.1 Pro is preserved here for reference.