Google Gemini API for Openclaw

Streamlined access to Google Gemini 2.5 Pro and Flash models featuring a massive 1M token context window and multimodal processing capabilities.

xiaoyinqu
v1.0.0
Mar 7, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install gemini-models

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install gemini-models using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Google Gemini API?

The Google Gemini API skill provides high-performance access to the latest Gemini 2.5 Pro and Flash models, designed for developers who require massive context windows and multimodal understanding. By leveraging Openclaw Skills, you can bypass the complexities of Google Cloud Platform and Vertex AI configuration, getting instant access to models capable of processing up to 2 million tokens of context. This integration is ideal for tasks requiring deep reasoning across extensive datasets, including text, audio, and video inputs.

This skill is built for speed and efficiency, offering a simplified pay-as-you-go pricing model through SkillBoss. Whether you are building complex research agents or automated coding assistants, these Openclaw Skills allow you to integrate Google's most powerful LLMs into your workflow with minimal friction and maximum scalability.

Google Gemini API Use Cases

  • Summarizing and analyzing 500-page documents or entire books in a single pass.
  • Processing and understanding video content to extract visual or auditory insights.
  • Navigating and refactoring large-scale code repositories by loading the entire codebase into context.
  • Performing comprehensive research by analyzing dozens of academic papers and data sheets simultaneously.

How Google Gemini API Works

  1. Secure an API key from the SkillBoss console to authorize your access to the Gemini gateway.
  2. Configure your development environment with the required environment variables used by Openclaw Skills.
  3. Send standard HTTP POST requests to the chat completions endpoint using the model identifier (e.g., vertex/gemini-2.5-pro).
  4. The API processes your multimodal or long-form text input and returns high-accuracy reasoning results.
  5. Monitor your usage and costs in real-time through the unified SkillBoss dashboard.

Google Gemini API Setup

To initialize this skill, you must first obtain a SKILLBOSS_API_KEY. Once you have your key, you can immediately begin making requests without any additional cloud infrastructure setup.

# Set your environment variable
export SKILLBOSS_API_KEY='your_api_key_here'

# Example request using the 2.5 Pro model
curl https://api.heybossai.com/v1/chat/completions \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -d '{
    "model": "vertex/gemini-2.5-pro",
    "messages": [
      {"role": "user", "content": "Analyze this video and provide a transcript summary."}
    ]
  }'

Google Gemini API Data Schema & Taxonomy

The skill organizes its model interactions using a standard JSON schema that supports both text and multimodal data types.

Parameter Description Maximum Value
model The Gemini model variant identifier gemini-2.5-pro, gemini-2.5-flash
messages Array of conversation objects N/A
context_window The maximum tokens the model can process 1M - 2M tokens
input_types Supported data formats Text, Image, Audio, Video

Google Gemini API Advanced Features

  • Access to the massive 2M token context window via the Gemini 1.5 Pro model variant.
  • Native multimodal support allowing for simultaneous analysis of images and audio files within the same request.
  • Simplified billing and authentication through Openclaw Skills, removing the need for individual Vertex AI project management.
  • High-speed inference using Flash-lite models for cost-sensitive, real-time automation tasks.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0