What would you like to be added?
I would like to request a new command or flag (for example: gemini models --output-format json or gemini --list-models) that outputs a list of all currently available Gemini models in a structured, machine-readable format like JSON.
The output should ideally include:
- Model ID (to be used with the -m flag)
- Display Name
- Short Description
- Context Limits (input/output token limits)
Why is this needed?
This feature is essential for developers building integrations or orchestrators around the Gemini CLI
Currently, third-party applications have to hardcode model names (e.g., gemini-1.5-pro, gemini-2.0-flash), which is problematic because:
- New models are released frequently, making hardcoded lists quickly outdated.
- Model availability varies based on the user's account, subscription plan, or region.
- The existing interactive /model command is great for human users in a REPL but cannot be programmatically parsed by external tools.
Having a machine-readable way to discover models will allow apps to dynamically populate selection menus and ensure users always have access to the latest models available to them.
Additional context
What would you like to be added?
I would like to request a new command or flag (for example: gemini models --output-format json or gemini --list-models) that outputs a list of all currently available Gemini models in a structured, machine-readable format like JSON.
The output should ideally include:
Why is this needed?
This feature is essential for developers building integrations or orchestrators around the Gemini CLI
Currently, third-party applications have to hardcode model names (e.g., gemini-1.5-pro, gemini-2.0-flash), which is problematic because:
Having a machine-readable way to discover models will allow apps to dynamically populate selection menus and ensure users always have access to the latest models available to them.
Additional context