Gemini CLI Update: Higher Limits for Google AI Pro and Ultra + New Data Cloud Extensions

September 25, 2025

3 min read

by Ufuk Ozen

Gemini CLI

Google AI Pro

Google AI Ultra

BigQuery

Cloud SQL

AlloyDB

Dataplex

Firestore

Looker

Spanner

MySQL

PostgreSQL

SQL Server

JSON Output Mode

Developer Tools

AI Agent

Google announces higher quota limits for Gemini CLI on Pro and Ultra, plus powerful extensions for AlloyDB, BigQuery, Cloud SQL, Dataplex, Firestore, Looker, and more.

If you're building with Gemini, today's update supercharges your workflow: Google AI Pro and Ultra subscribers now get significantly higher request limits for Gemini CLI, giving you more headroom to ship faster with Gemini 2.5 Pro and Flash.

What’s new: higher limits for Gemini CLI

Google is rolling out higher model request limits for Gemini CLI to Google AI Pro and Ultra subscribers, enabling longer sessions and more complex tasks without hitting quotas. This is especially valuable for IDE-mode workflows, terminal-first development, and automated CI.

Connect all your data: Gemini CLI extensions for Google Data Cloud

Google introduced open-source Gemini CLI extensions for Google Data Cloud services, letting you manage and analyze data directly from your terminal. Supported services include:

AlloyDB (interact, manage, and observe)
BigQuery (data analytics and conversational insights)
Cloud SQL for PostgreSQL, MySQL, SQL Server (manage, observe)
Dataplex (discover, manage, and govern data + AI artifacts)
Firestore (interact with databases, collections, and documents)
Looker (query data, run Looks, create dashboards)
Spanner (interact with databases and data)
MySQL and Postgres (generic database tools)
SQL Server (generic database tools)
MCP Toolbox (load custom tools across 30+ data sources)

Quick start (step-by-step)

Install the latest Gemini CLI:


1L • 4W
example.sh
1npm install -g @google/gemini-cli@latest
BASH1 lines40 chars
UTF-8

Install an extension (example: Cloud SQL for PostgreSQL):


1L • 4W
example.sh
1gemini extensions install https://github.com/gemini-cli-extensions/cloud-sql-postgresql
BASH1 lines87 chars
UTF-8

Configure environment variables for your Google Cloud project and service credentials.
Launch the CLI and list installed extensions:


3L • 4W
example.sh
1gemini
2/extensions
3/mcp list
BASH3 lines28 chars
UTF-8

Provision and manage resources with natural-language prompts (e.g., create instances, users, and tables), then generate APIs and seed test data for faster integration.

Power features that boost productivity

Alongside the extensions and higher limits, Gemini CLI adds several developer-focused enhancements:

JSON Output Mode: Run headlessly with --output-format json to get structured responses for automated parsing, logging, or CI workflows.
Keybinding Triggered Approvals: When YOLO/auto-edit modes are activated using shortcuts (shift+y or shift+tab), pending confirmations will auto-approve.
Chat Sharing: Export the current conversation as Markdown or JSON with /chat share file.md|file.json.
Prompt Search: Quickly search your prompt history using ctrl+r.
Input Undo/Redo: Recover edited content with ctrl+z (undo) and ctrl+shift+z (redo).
Loop Detection Confirmation: Disable loop detection per session if needed via a confirmation dialog.
Direct to Google Cloud Telemetry: Streamlined telemetry pipeline to Google Cloud.
Visual Mode Indicator Revamp: Clear color-coded indicators for shell, accept edits, and yolo modes.

Why this matters for teams

Faster iteration: Higher limits mean less friction when running long sessions, multi-step tasks, or CI-driven workflows.
Data-in-the-loop: Extensions let you build, query, and monitor databases without context switching—right from your terminal.
Production-ready automation: JSON output and approvals help you wire Gemini into pipelines with strong guarantees.