> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mountsea.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

# Introduction to Google Gemini

Welcome to the **Google Gemini** service documentation!

Powered by Google's cutting-edge AI, this service provides three major capabilities organized into dedicated sections:

<CardGroup cols={3}>
  <Card title="Video Generation" icon="video" href="/api-reference/gemini/video-introduction">
    Generate and edit videos with **Veo 2 / Veo 3 / Veo 3.1** — text-to-video, image-to-video, extend, reshoot, object insert/remove
  </Card>

  <Card title="Image Generation" icon="image" href="/api-reference/gemini/image-introduction">
    Create and edit images with **Nano Banana** (Fast / Pro / 2) — supports multiple aspect ratios and up to 4K resolution
  </Card>

  <Card title="Gemini Compat (SDK)" icon="plug" href="/api-reference/gemini/compat-introduction">
    Drop-in replacement for Google's official **@google/genai** SDK — just change the base URL
  </Card>
</CardGroup>

## Service Overview

### 🎬 Video Generation (Veo 2 / 3 / 3.1)

Full-featured video generation and editing pipeline:

* **Generate Videos** — Text-to-video, image-to-video (start/end frame), ingredients-to-video (1-3 reference images)
* **Upsample** — Upscale to **1080p** or **4K**, or generate **GIF** previews
* **Extend** — Extend video duration seamlessly (Veo 3.1 only)
* **Reshoot** — Re-generate with different camera motion (pan, zoom, dolly, etc.)
* **Object Insert/Remove** — Add or remove objects using masks
* **Prompt Expand** — Expand simple prompts into cinematic descriptions

[Explore Video Documentation →](/api-reference/gemini/video-introduction)

***

### 🖼️ Image Generation (Nano Banana)

Powerful image generation and editing with three model tiers:

* **Text-to-Image** — Generate original images from text prompts
* **Image Editing** — Modify existing images based on prompts
* **Multiple Aspect Ratios** — 21:9, 16:9, 9:16, 4:3, 3:2, 5:4, 1:1, and more
* **High Resolution** — Up to **4K** output (Pro / 2 models)

Available models: `nano-banana-fast`, `nano-banana-pro`, `nano-banana-2`

[Explore Image Documentation →](/api-reference/gemini/image-introduction)

***

### 🔌 Gemini Compat (Official SDK)

A **Google Gemini API-compatible** endpoint that works with Google's official **@google/genai** SDK:

* **Drop-in replacement** — Just change the base URL to `https://api.mountsea.ai/gemini`
* **Text generation** — `generateContent` and `streamGenerateContent`
* **Image generation** — Use Gemini model IDs (`gemini-2.5-flash-image`, `gemini-3.1-flash-image-preview`, `gemini-3-pro-image-preview`) — automatically routed to Nano Banana models
* **Supports official SDK** — Use the TypeScript/Python `@google/genai` library directly

[Explore Compat Documentation →](/api-reference/gemini/compat-introduction)

***

## Base URL

All Gemini endpoints are accessed at:

```
https://api.mountsea.ai
```

For the Gemini Compat SDK endpoint, set your SDK base URL to:

```
https://api.mountsea.ai/gemini
```

## Authentication

All endpoints require Bearer token authentication:

```
Authorization: Bearer your-api-key
```

## Shared Task Polling

Video and Image async APIs share one polling endpoint:

| Endpoint              | Method | Description                                     |
| --------------------- | ------ | ----------------------------------------------- |
| `/gemini/task/result` | GET    | Poll any Gemini video or image task by `taskId` |

See [Get Task Result](/api-reference/gemini/task).