Google Introduces Gemini CLI, an Open-Source AI Tool for Developers
Google is rolling out a new AI-powered tool to bring its Gemini AI models directly into developers' coding environments.
On Wednesday, Google unveiled Gemini CLI, an AI-driven tool crafted to operate locally from a terminal. This innovative tool integrates Google's Gemini AI models with local codebases, enabling developers to use natural language to request explanations for complex code, create new features, debug issues, or execute commands.
Gemini CLI is a key part of Google’s strategy to embed its AI models into developers’ coding routines. Alongside tools like Gemini Code Assist and the asynchronous AI coding assistant Jules, Google is expanding its AI coding portfolio. However, Gemini CLI directly rivals command-line AI tools like OpenAI’s Codex CLI and Anthropic’s Claude Code, which are often praised for their seamless integration, speed, and efficiency.
Since the launch of Gemini 2.5 Pro in April, Google’s AI models have gained significant traction among developers. The success of Gemini 2.5 Pro has fueled the adoption of third-party AI coding tools like Cursor and GitHub Copilot, which have grown into major platforms. In response, Google has focused on fostering direct connections with developers by introducing its own in-house solutions.
Beyond coding, Gemini CLI is designed for versatility. Developers can leverage it to generate videos using Google’s Veo 3 model, produce research reports with the Deep Research agent, or retrieve real-time data via Google Search. Additionally, Gemini CLI supports connections to MCP servers, allowing integration with external databases.
To drive adoption, Google is releasing Gemini CLI as open-source software under the Apache 2.0 license, known for its permissive terms. The company anticipates active contributions from the developer community on GitHub.
Google is also offering substantial usage limits to encourage adoption of Gemini CLI. Free users can make up to 60 model requests per minute and 1,000 requests daily, roughly twice the average usage Google observed during testing.
Save Over $200 on Your TechCrunch All Stage Pass
Innovate smarter. Scale quicker. Connect meaningfully. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day of insights, workshops, and valuable networking.
I notice that the HTML content you provided includes some incomplete or misplaced elements at the end, such as "Techcrunch event," "Boston, MA," "July 15," and "REGISTER NOW," which appear to be out of context and not properly wrapped in HTML tags. Additionally, there’s an empty tag. For the purpose of rewriting, I’ll focus on the main content and the properly structured HTML elements, as per the rules, and exclude the incomplete or unclear elements to maintain the integrity of the HTML structure. If you want those elements included, please provide additional context or clarification.
Below is the rewritten content following your instructions, with the title and HTML content properly formatted and rewritten in natural, SEO-compliant English, preserving all HTML tags and their structure.
Related article
Google Expands NotebookLM Access to Younger Students Amid Rising AI Education Competition
Google’s AI-powered note-taking tool is now accessible to younger users, previously restricted to those 18 and older. The company announced that NotebookLM is available for Google Workspace for Educat
Google Unveils A2A Protocol to Enhance AI Agent Interoperability
AI agents tackle complex, repetitive tasks like supply chain management and equipment procurement. As organizations adopt agents from various vendors and frameworks, these agents often become isolated
ByteDance Unveils Seed-Thinking-v1.5 AI Model to Boost Reasoning Capabilities
The race for advanced reasoning AI began with OpenAI’s o1 model in September 2024, gaining momentum with DeepSeek’s R1 launch in January 2025.Major AI developers are now competing to create faster, mo
Comments (0)
0/200
Google is rolling out a new AI-powered tool to bring its Gemini AI models directly into developers' coding environments.
On Wednesday, Google unveiled Gemini CLI, an AI-driven tool crafted to operate locally from a terminal. This innovative tool integrates Google's Gemini AI models with local codebases, enabling developers to use natural language to request explanations for complex code, create new features, debug issues, or execute commands.
Gemini CLI is a key part of Google’s strategy to embed its AI models into developers’ coding routines. Alongside tools like Gemini Code Assist and the asynchronous AI coding assistant Jules, Google is expanding its AI coding portfolio. However, Gemini CLI directly rivals command-line AI tools like OpenAI’s Codex CLI and Anthropic’s Claude Code, which are often praised for their seamless integration, speed, and efficiency.
Since the launch of Gemini 2.5 Pro in April, Google’s AI models have gained significant traction among developers. The success of Gemini 2.5 Pro has fueled the adoption of third-party AI coding tools like Cursor and GitHub Copilot, which have grown into major platforms. In response, Google has focused on fostering direct connections with developers by introducing its own in-house solutions.
Beyond coding, Gemini CLI is designed for versatility. Developers can leverage it to generate videos using Google’s Veo 3 model, produce research reports with the Deep Research agent, or retrieve real-time data via Google Search. Additionally, Gemini CLI supports connections to MCP servers, allowing integration with external databases.
To drive adoption, Google is releasing Gemini CLI as open-source software under the Apache 2.0 license, known for its permissive terms. The company anticipates active contributions from the developer community on GitHub.
Google is also offering substantial usage limits to encourage adoption of Gemini CLI. Free users can make up to 60 model requests per minute and 1,000 requests daily, roughly twice the average usage Google observed during testing.
Save Over $200 on Your TechCrunch All Stage Pass
Innovate smarter. Scale quicker. Connect meaningfully. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day of insights, workshops, and valuable networking.
I notice that the HTML content you provided includes some incomplete or misplaced elements at the end, such as "Techcrunch event," "Boston, MA," "July 15," and "REGISTER NOW," which appear to be out of context and not properly wrapped in HTML tags. Additionally, there’s an empty tag. For the purpose of rewriting, I’ll focus on the main content and the properly structured HTML elements, as per the rules, and exclude the incomplete or unclear elements to maintain the integrity of the HTML structure. If you want those elements included, please provide additional context or clarification.
Below is the rewritten content following your instructions, with the title and HTML content properly formatted and rewritten in natural, SEO-compliant English, preserving all HTML tags and their structure.
Google Expands NotebookLM Access to Younger Students Amid Rising AI Education Competition
Google’s AI-powered note-taking tool is now accessible to younger users, previously restricted to those 18 and older. The company announced that NotebookLM is available for Google Workspace for Educat
Google Unveils A2A Protocol to Enhance AI Agent Interoperability
AI agents tackle complex, repetitive tasks like supply chain management and equipment procurement. As organizations adopt agents from various vendors and frameworks, these agents often become isolated
ByteDance Unveils Seed-Thinking-v1.5 AI Model to Boost Reasoning Capabilities
The race for advanced reasoning AI began with OpenAI’s o1 model in September 2024, gaining momentum with DeepSeek’s R1 launch in January 2025.Major AI developers are now competing to create faster, mo
