Google Intros Open Source Gemini Model For Developers

Gemini CLI (command line interface), announced Wednesday (June 25), is an open-source AI agent that offers “lightweight” access to the company’s Gemini model. Despite its name, however, Google says the tool is designed for coding as well as tasks like content generation, deep research and task management.

“For developers, the command line interface isn’t just a tool; it’s home,” the company said in its announcement. “The terminal’s efficiency, ubiquity and portability make it the go-to utility for getting work done. And as developers’ reliance on the terminal endures, so does the demand for integrated AI assistance.”

Google said it has also integrated Gemini CLI with its AI coding assistant, Gemini Code Assist, allowing all developers — on free, Standard and Enterprise Code Assist plans — to “get prompt-driven, AI-first coding in both VS Code and Gemini CLI.”

Also this week, Google’s DeepMind introduced a vision language action (VLA) model that runs locally on robotic devices, with no need to access a data network.

The new robotics foundation model, Gemini Robotics On-Device, features general-purpose dexterity and fast task adaptation, the company wrote on its blog.

“Since the model operates independent of a data network, it’s helpful for latency sensitive applications and ensures robustness in environments with intermittent or zero connectivity,” Google DeepMind Senior Director and Head of Robotics Carolina Parada said in the post.

The model follows natural language instructions and has the dexterity to carry out tasks like unzipping bags, folding clothes, zipping a lunchbox, drawing a card, pouring salad dressing and assembling products, and is also Google DeepMind’s first VLA (vision-language-action) model that is available for fine-tuning, per the post.

In other AI news, PYMNTS wrote Wednesday about the technology’s use in the B2B payments space as that industry adopts agentic AI.

“I think we’re standing at the edge of a major transformation,” Coupa Product Strategy and Management SVP Rajiv Ramachandran told PYMNTS during an interview for the June “What’s Next in Payments: Secret Agents” series.

“Agentic AI is not just a technological trend — it’s a rethinking of how decision-making, risk and value creation happen inside financial workflows,” he added.

While “agentic AI” may sound futuristic, Ramachandran said he sees it as a pragmatic solution to an old business imperative: do more, faster and more securely, with less human overhead.

“It’s not just about automation; it’s about delegation,” he said.