llm-gemini Plugin Adds Support for Google's Gemini 3.5 Flash Model

Developer Simon Willison released version 0.32 of the llm-gemini plugin on May 19, 2026, adding support for Google's recently announced Gemini 3.5 Flash model.

The llm-gemini plugin provides command-line access to Google's Gemini model family through Willison's LLM tool, which offers a unified interface for interacting with multiple large language models from the terminal.

What Changed

Version 0.32 introduces the new gemini-3.5-flash model identifier, allowing developers to access Google's latest Flash-tier model directly from the command line. The plugin previously supported earlier versions of Google's Gemini models.

Willison demonstrated the plugin's functionality by using it to generate a pelican drawing with Gemini 3.5 Flash, according to his blog post.

Technical Details

The LLM tool and its plugins are built in Python and designed for developers who prefer command-line interfaces for AI model interaction. The llm-gemini plugin specifically handles authentication and API communication with Google's Gemini service.

Pricing and specific technical specifications for Gemini 3.5 Flash through the plugin were not disclosed in the release notes.

What This Means

This incremental update maintains the llm-gemini plugin's position as a command-line option for developers working with Google's Gemini models. The quick integration of Gemini 3.5 Flash indicates active maintenance of the plugin, though the update itself is routine rather than introducing new functionality beyond model support. For developers already using the LLM ecosystem, this provides immediate access to Google's latest Flash-tier model without changing workflows.