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

# VEO Video Generation API (Custom API - Deprecated)

> AI video generation service based on VEO3 series models - Legacy custom API, not recommended

**⚠️ This API is Deprecated, Not Recommended**Please use the new [Veo-3.1 API](/en/api-capabilities/veo/veo-31-overview), which provides:

* ✅ OpenAI-compatible format, easier integration
* ✅ Lower prices (0.15−0.15-0.15−0.25/request)
* ✅ More model options (8 variants)
* ✅ Image-to-video functionality

**Legacy VEO3 custom API is maintained for compatibility only, no new features will be added.**

**Base URL:** `https://api.yelinai.com`
**API Version:** `v1`
**Service Status:** Maintenance mode (Not recommended for new projects)

## VEO vs Sora 2 Quick Comparison

**Important Decision:** Choosing the right tool can save you up to **90%** in costs!

## VEO - Professional Quality First

**Price:** 2−2-2−10/call**Core Advantages:**

* 🎨 Supports up to 5 reference images
* 🤖 AI-powered prompt enhancement
* 👑 Professional-grade quality (Pro version)
* 🎬 Frame-level precise control

**Best For:**

* Commercial advertising
* Film post-production effects
* Brand promotional videos
* Multi-image reference needs

## Sora 2 - Cost-Effective

**Price:** \$0.15/call (13-67x cheaper)**Core Advantages:**

* ⚡ Fast generation (2-3 mins)
* 💰 Extremely low cost
* 🎵 Audio-video sync
* 🔥 Ultra-high stability

**Best For:**

* Social media content
* Rapid testing and iteration
* Batch video generation
* Budget-conscious projects

**Money-Saving Tip:** Test with Sora 2 (0.15)for5−10promptsfirst,thenuseVEO(0.15) for 5-10 prompts first, then use VEO (0.15)for5−10promptsfirst,thenuseVEO(2-\$10) for final production. Save **70-90%** on testing costs<img src="https://mintlify.s3.us-west-1.amazonaws.com/yelinkeji/en/api-capabilities/sora2/overview" alt="View Sora 2 Docs →" />

## Overview

VEO API provides powerful AI video generation capabilities, supporting 4 different VEO3 series models. The service uses asynchronous processing mode to ensure efficient handling of video generation requests.

## Core Features

## Async Processing

Instant response after task submission, retrieve results via polling

## Multi-Model Support

4 VEO3 models for different needs

## Image Reference

Supports multiple reference images for generation

## Prompt Enhancement

AI optimizes prompts for better results

## Pay on Success

Only charged for successfully generated videos

## High-Quality Output

Supports HD video generation, 10-15 seconds duration

## Workflow

1

Submit Task

Call submit endpoint with text description and parameters

2

Get Task ID

System returns unique task ID

3

Poll Status

Use task ID to query processing progress

4

Get Result

Obtain video URL after task completion

## Supported Models

| Model             | Price   | Features                                        | Generation Time |
| ----------------- | ------- | ----------------------------------------------- | --------------- |
| `veo3`            | \$2.00  | Standard version, excellent overall performance | 3-8 minutes     |
| `veo3-fast`       | \$2.00  | Fast version, audio support                     | 1-3 minutes     |
| `veo3-pro`        | \$10.00 | Professional version, exceptional quality       | 5-15 minutes    |
| `veo3-pro-frames` | \$10.00 | Pro version + frame control                     | 5-15 minutes    |

## Getting Started

## Get API Key

Contact **[threezhang.cn@gmail.com](mailto:threezhang.cn@gmail.com)** or WeChat **Kikivivikids** to get API key

## Related Links

## Quick Start

Learn how to get started quickly

## API Reference

View detailed API documentation

## Code Examples

View examples in various programming languages

## Best Practices

Learn how to achieve best generation results