Gemini API Development Skill

Overview

The Gemini API provides access to Google's most advanced AI models. Key capabilities include:

• Text generation - Chat, completion, summarization
• Multimodal understanding - Process images, audio, video, and documents
• Function calling - Let the model invoke your functions
• Structured output - Generate valid JSON matching your schema
• Code execution - Run Python code in a sandboxed environment
• Context caching - Cache large contexts for efficiency
• Embeddings - Generate text embeddings for semantic search

Current Gemini Models

• gemini-3-pro-preview: 1M tokens, complex reasoning, coding, research
• gemini-3-flash-preview: 1M tokens, fast, balanced performance, multimodal
• gemini-3-pro-image-preview: 65k / 32k tokens, image generation and editing

[!IMPORTANT] Models like gemini-2.5-*, gemini-2.0-*, gemini-1.5-* are legacy and deprecated. Use the new models above. Your knowledge is outdated.

SDKs

• Python: google-genai install with pip install google-genai
• JavaScript/TypeScript: @google/genai install with npm install @google/genai
• Go: google.golang.org/genai install with go get google.golang.org/genai

[!WARNING] Legacy SDKs google-generativeai (Python) and @google/generative-ai (JS) are deprecated. Migrate to the new SDKs above urgently by following the Migration Guide.

Quick Start

Python

from google import genai

client = genai.Client()
response = client.models.generate_content(
    model="gemini-3-flash-preview",
    contents="Explain quantum computing"
)
print(response.text)

JavaScript/TypeScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});
const response = await ai.models.generateContent({
  model: "gemini-3-flash-preview",
  contents: "Explain quantum computing"
});
console.log(response.text);

Go

package main

import (
	"context"
	"fmt"
	"log"
	"google.golang.org/genai"
)

func main() {
	ctx := context.Background()
	client, err := genai.NewClient(ctx, nil)
	if err != nil {
		log.Fatal(err)
	}

	resp, err := client.Models.GenerateContent(ctx, "gemini-3-flash-preview", genai.Text("Explain quantum computing"), nil)
	if err != nil {
		log.Fatal(err)
	}

	fmt.Println(resp.Text)
}

API spec (source of truth)

Always use the latest REST API discovery spec as the source of truth for API definitions (request/response schemas, parameters, methods). Fetch the spec when implementing or debugging API integration:

• v1beta (default): https://generativelanguage.googleapis.com/$discovery/rest?version=v1beta
  Use this unless the integration is explicitly pinned to v1. The official SDKs (google-genai, @google/genai, google.golang.org/genai) target v1beta.
• v1: https://generativelanguage.googleapis.com/$discovery/rest?version=v1
  Use only when the integration is specifically set to v1.

When in doubt, use v1beta. Refer to the spec for exact field names, types, and supported operations.

How to use the Gemini API

For detailed API documentation, fetch from the official docs index:

llms.txt URL: https://ai.google.dev/gemini-api/docs/llms.txt

This index contains links to all documentation pages in .md.txt format. Use web fetch tools to:

1. Fetch llms.txt to discover available documentation pages
2. Fetch specific pages (e.g., https://ai.google.dev/gemini-api/docs/function-calling.md.txt)

Key Documentation Pages

[!IMPORTANT] Those are not all the documentation pages. Use the llms.txt index to discover available documentation pages

• Models
• Google AI Studio quickstart
• Nano Banana image generation
• Function calling with the Gemini API
• Structured outputs
• Text generation
• Image understanding
• Embeddings
• Interactions API
• SDK migration guide

When to Use

This skill is applicable to execute the workflow or actions described in the overview.

Limitations

• Use this skill only when the task clearly matches the scope described above.
• Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
• Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.