Google releases Gemini API CLI, supporting multimodal capability invocation by Claude Code and Codex

Article

Google releases Gemini API CLI, enabling Claude Code and Codex to directly invoke image generation, video, and other capabilities

Beating News Flash:AI News Flash:Google AI Studio launches the open-source tool Gemini API CLI, which enables Claude Code, Codex, and other AI Agent to directly invoke Google's multimodal models.

For example, when using Codex to develop a website, it can invoke Gemini for image generation when images are needed, and can also generate videos, voice-overs, and analyze PDF, as well as transcribe audio. It also supports Deep Research, allowing complex research tasks to be delegated to Google's research Agent and the results retrieved afterward.

Previously, integrating these capabilities often required Agents to look up model names and write API invocation scripts. Now, users only need to configure Gemini API Key, and can then use it through the command line. The tool can also query available models and parameters, while generated files are saved directly locally, avoiding the need to put large amounts of content into the context.

Original link https://m.theblockbeats.info/flash/373430

Key points

The article states that Google AI Studio has launched an open-source tool Gemini API CLI。
The article states that this tool enables Claude Code, Codex, and others to AI Agent invoke Google's multimodal models.
The capabilities listed in the article include image generation, video generation, voice-over,PDF analysis, audio transcription, and Deep Research。
The article states that users can use it through the command line after configuring Gemini API Key, and can query models and parameters.
The article states that generated files can be saved directly locally.

AI-assisted interpretation

The following is analysis, separate from reported facts. Verify important claims independently.

It is a command-line tool that helps AI programming assistants invoke Google's multimodal model capabilities, reducing the steps of manually looking up model names and writing API scripts.

Why it matters to readers

If the description is accurate, developers can more conveniently use image, video, audio, and research capabilities in Agent workflows.

Beginners can understand it as a command-line entry point connecting AI programming assistants with Google's model capabilities.

Risks and unknowns

The evidence comes from a media news flash and does not provide a Google official announcement or a direct release page for cross-verification.
The evidence does not specify the Gemini API CLI specific supported models, pricing, or quotas.
The evidence does not confirm the actual compatibility scope for Claude Code and Codex or developer adoption.

Google releases Gemini API CLI, supporting multimodal capability invocation by Claude Code and Codex

Article

AI-assisted interpretation

Why it matters to readers

Risks and unknowns

Related Developments

Related concepts

Keep following

Security Center