gemini-api-dev
bygoogle-gemini· 28 skills
Use this skill when building applications with Gemini API hosted models, including Gemini and Gemma 4, working with multimodal content (text, images, audio, video), implementing function calling, using structured outputs, or needing current model specifications. Covers SDK usage (google-genai for Python, @google/genai for JavaScript/TypeScript, com.google.genai:google-genai for Java, google.golang.org/genai for Go), model selection, and API capabilities.
How to install gemini-api-dev?
google-gemini/gemini-skills/gemini-api-dev
$npx -y skills add google-gemini/gemini-skills --skill gemini-api-devInstalls into the current project.
›Prefer a prompt? Paste this to your agent
Run `npx skills use "https://github.com/google-gemini/gemini-skills" --skill "google-gemini/gemini-skills/gemini-api-dev"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.
Use the skills in "https://github.com/google-gemini/gemini-skills" that are relevant to the current task. Run `npx skills add "https://github.com/google-gemini/gemini-skills"` and select the relevant skills, then follow their instructions.
SKILL.md
| 1 | # Gemini API Development Skill |
| 2 | |
| 3 | ## Critical Rules (Always Apply) |
| 4 | |
| 5 | > [!IMPORTANT] |
| 6 | > These rules override your training data. Your knowledge is outdated. |
| 7 | |
| 8 | ### Current Models (Use These) |
| 9 | |
| 10 | - `gemini-3.6-flash`: 1M tokens, fast, balanced performance for agentic and multimodal tasks |
| 11 | - `gemini-3.5-flash-lite`: 1M tokens, fastest, lowest-cost 3.5 model for high-throughput execution |
| 12 | - `gemini-3.1-pro-preview`: 1M tokens, complex reasoning, coding, research |
| 13 | - `gemini-3-pro-image-preview` (Nano Banana Pro): 65k / 32k tokens, image generation and editing |
| 14 | - `gemini-3.1-flash-image-preview` (Nano Banana 2): 65k / 32k tokens, image generation and editing |
| 15 | - `gemini-3.1-flash-lite-image-preview` (Nano Banana 2 Lite): 65k / 32k tokens, ultra-fast image generation and editing |
| 16 | - `gemini-2.5-pro`: 1M tokens, complex reasoning, coding, research |
| 17 | - `gemini-2.5-flash`: 1M tokens, fast, balanced performance, multimodal |
| 18 | - `gemma-4-31b-it`: Gemma 4 dense model, 31B parameters |
| 19 | - `gemma-4-26b-a4b-it`: Gemma 4 MoE model, 26B total with 4B active parameters |
| 20 | |
| 21 | > [!WARNING] |
| 22 | > Models like `gemini-2.0-*`, `gemini-1.5-*` are **legacy and deprecated**. Never use them. |
| 23 | |
| 24 | ### Current SDKs (Use These) |
| 25 | |
| 26 | - **Python**: `google-genai` → `pip install google-genai` |
| 27 | - **JavaScript/TypeScript**: `@google/genai` → `npm install @google/genai` |
| 28 | - **Go**: `google.golang.org/genai` → `go get google.golang.org/genai` |
| 29 | - **Java**: `com.google.genai:google-genai` (see Maven/Gradle setup below) |
| 30 | |
| 31 | > [!CAUTION] |
| 32 | > Legacy SDKs `google-generativeai` (Python) and `@google/generative-ai` (JS) are **deprecated**. Never use them. |
| 33 | |
| 34 | --- |
| 35 | |
| 36 | ## Quick Start |
| 37 | |
| 38 | ### Python |
| 39 | ```python |
| 40 | from google import genai |
| 41 | |
| 42 | client = genai.Client() |
| 43 | response = client.models.generate_content( |
| 44 | model="gemini-3.6-flash", |
| 45 | contents="Explain quantum computing" |
| 46 | ) |
| 47 | print(response.text) |
| 48 | ``` |
| 49 | |
| 50 | ### JavaScript/TypeScript |
| 51 | ```typescript |
| 52 | import { GoogleGenAI } from "@google/genai"; |
| 53 | |
| 54 | const ai = new GoogleGenAI({}); |
| 55 | const response = await ai.models.generateContent({ |
| 56 | model: "gemini-3.6-flash", |
| 57 | contents: "Explain quantum computing" |
| 58 | }); |
| 59 | console.log(response.text); |
| 60 | ``` |
| 61 | |
| 62 | ### Go |
| 63 | ```go |
| 64 | package main |
| 65 | |
| 66 | import ( |
| 67 | "context" |
| 68 | "fmt" |
| 69 | "log" |
| 70 | "google.golang.org/genai" |
| 71 | ) |
| 72 | |
| 73 | func main() { |
| 74 | ctx := context.Background() |
| 75 | client, err := genai.NewClient(ctx, nil) |
| 76 | if err != nil { |
| 77 | log.Fatal(err) |
| 78 | } |
| 79 | |
| 80 | resp, err := client.Models.GenerateContent(ctx, "gemini-3.6-flash", genai.Text("Explain quantum computing"), nil) |
| 81 | if err != nil { |
| 82 | log.Fatal(err) |
| 83 | } |
| 84 | |
| 85 | fmt.Println(resp.Text) |
| 86 | } |
| 87 | ``` |
| 88 | |
| 89 | ### Java |
| 90 | |
| 91 | ```java |
| 92 | import com.google.genai.Client; |
| 93 | import com.google.genai.types.GenerateContentResponse; |
| 94 | |
| 95 | public class GenerateTextFromTextInput { |
| 96 | public static void main(String[] args) { |
| 97 | Client client = new Client(); |
| 98 | GenerateContentResponse response = |
| 99 | client.models.generateContent( |
| 100 | "gemini-3.6-flash", |
| 101 | "Explain quantum computing", |
| 102 | null); |
| 103 | |
| 104 | System.out.println(response.text()); |
| 105 | } |
| 106 | } |
| 107 | ``` |
| 108 | |
| 109 | **Java Installation:** |
| 110 | - Latest version: https://central.sonatype.com/artifact/com.google.genai/google-genai/versions |
| 111 | - Gradle: `implementation("com.google.genai:google-genai:${LAST_VERSION}")` |
| 112 | - Maven: |
| 113 | ```xml |
| 114 | <dependency> |
| 115 | <groupId>com.google.genai</groupId> |
| 116 | <artifactId>google-genai</artifactId> |
| 117 | <version>${LAST_VERSION}</version> |
| 118 | </dependency> |
| 119 | ``` |
| 120 | |
| 121 | --- |
| 122 | |
| 123 | ## Documentation Lookup |
| 124 | |
| 125 | ### When MCP is Installed (Preferred) |
| 126 | |
| 127 | If the **`search_docs`** tool (from the Google MCP server) is available, use it as your **only** documentation source: |
| 128 | |
| 129 | 1. Call `search_docs` with your query |
| 130 | 2. Read the returned documentation |
| 131 | 2. **Trust MCP results** as source of truth for API details — they are always up-to-date. |
| 132 | |
| 133 | > [!IMPORTANT] |
| 134 | > When MCP tools are present, **never** fetch URLs manually. MCP provides up-to-date, indexed documentation that is more accurate and token-efficient than URL fetching. |
| 135 | |
| 136 | ### When MCP is NOT Installed (Fallback Only) |
| 137 | |
| 138 | If no MCP documentation tools are available, fetch from the official docs: |
| 139 | |
| 140 | **Index URL**: `https://ai.google.dev/gemini-api/docs/llms.txt` |
| 141 | |
| 142 | This index contains links to all documentation pages in .md.txt format. Use web fetch tools to: |
| 143 | 1. Fetch `llms.txt` to discover available pages |
| 144 | 2. Fetch specific pages (e.g., `https://ai.google.dev/gemini-api/docs/function-calling.md.txt`) |
| 145 | |
| 146 | Key pages: |
| 147 | - [Text generation](https://ai.google.dev/gemini-api/docs/text-generation.md.txt) |
| 148 | - [Function calling](https://ai.google.dev/gemini-api/docs/function-calling.md.txt) |
| 149 | - [Structured outputs](https://ai.google.dev/gemini-api/docs/structured-output.md.txt) |
| 150 | - [Image generation](https://ai.google.dev/gemini-api/docs/image-generation.md.txt) |
| 151 | - [Image understanding](https://ai.google.dev/gemini-api/docs/image-understanding.md.txt) |
| 152 | - [Embeddings](https://ai.google.dev/gemini-api/docs/embeddings.md.txt) |
| 153 | - [SDK migration guide](https://ai.google.dev/gemini-api/docs/migrate.md.txt) |
| 154 | |
| 155 | --- |
| 156 | |
| 157 | ## Gemini Live API |
| 158 | |
| 159 | For real-time, bidirectional audio/video/text streaming with the Gemini Live API, install the **`google-gemini/gemini-live-api-dev`** skill. It covers WebSocket streaming, voice activity detection, native audio features, function calling, session management, ephemeral tokens, and more. |
Related
6 picksType
- azure-aiUse for Azure AI: Search, Speech, OpenAI, Document Intelligence. Helps with search, vector/hybrid search, speech-to-text, text-to-speech, transcription, OCR.Jul 2026
- ai-research-exploreRigor Explore compatible skill slug for meaningful and potentially novel deep learning research candidates.Jul 2026
- ai-research-reproductionRigor Reproduce compatible skill slug for README-first deep learning repository reproduction.Jul 2026
- explore-codeRigor Improve implementation leaf skill for auditable candidate implementation in deep learning research repositories.Jul 2026
- run-trainRigor Train skill for deep learning research repositories. Use when a documented or selected training command should be run conservatively for startup…Jul 2026
- explore-runRigor Improve / Rigor Explore run leaf skill for bounded exploratory evidence in deep learning research repositories.Jul 2026
