Sora 2 Video Generation API

Base URL: https://api.yelinai.com/v1/chat/completions API Version: v1 Service Status: Running | No Watermark | No Invitation Code Required Need ultimate stability? If the current solution doesn’t meet your stability requirements, try the Official Forward Plan (OpenAI official API transparent forwarding, 99.99% availability).

​

Prerequisites

1 Get API Key Log in to laozhang.ai console to obtain your API key 2 Configure Billing Mode Edit token settings and choose one of the following billing modes (same price for both):

• Volume Priority (Recommended): Uses balance billing first, automatically switches when balance is insufficient. Suitable for most users
• Pay-per-call: Direct deduction for each call. Suitable for strict budget control scenarios

Both modes have exactly the same price at $0.15/call (10s or 15s), only the billing method differs. If billing mode is not configured, API calls will fail. You must complete this configuration first!

​

Major Updates

Release Date: October 1, 2025 OpenAI suddenly released Sora 2 new version (The New Sora), and we integrated official website API immediately (reverse-engineered, before OpenAI released official API). Latest Updates

• 01/04 New Character Creation API, extract characters from videos and reuse via @username ($0.01/call)
• 10/20 Massive price drop for 15s models! From 0.25to0.25 to 0.25to0.15/call, all models unified pricing; Added raw error log feature
• 10/18 Added support for generating videos with Sora native watermark (default is no watermark)
• 10/16 Restored 15-second models (sora_video2-15s and sora_video2-landscape-15s)
• 10/10 Added sora-2-pro HD model, supports HD 1080P and 15-second duration (Async API only)
• 10/07 Added Python example code; official sora-2 and sora-2-pro models launched but with high pricing and rate limits
• 10/01 Integrated Sora 2 reverse-engineered API, supports text-to-video and image-to-video

​

Sora 2 Core Breakthroughs

​

Audio-Video Sync

First-time synchronized audio-video generation

​

Physical Realism

Dramatically improved physical realism

​

Long Video Support

Supports 10-16 second coherent narrative videos

​

No Watermark Output

Generated videos have no watermark (official website has watermarks)

​

Cameo Feature

Supports integrating real-person recorded clips into AI scenes

​

Remix Feature

Supports community remixing

​

Core Features

​

Text-to-Video

Generate videos from pure text descriptions

​

Image-to-Video

Supports reference images (up to 1 image)

​

Character Creation

Extract characters from videos, reuse via @username

​

Async API

More stable task queue approach (recommended for production)

​

Multi-Spec Support

Horizontal and vertical orientation options

​

Pay-per-Use

Charged only by generation count

​

Workflow

1 Configure API Key Get API key from laozhang.ai console, set to pay-as-you-go or pay-per-use mode 2 Send Request Call /v1/chat/completions endpoint, specify model and prompt 3 Real-time Progress View task queue and generation progress through streaming output 4 Get Video Obtain video link after generation completes (valid for 1 day) 5 Download & Save Download video to local storage immediately **Price Update (10/20)**🎉 15-second models price drop! From 0.25to0.25 to 0.25to0.15/callAll base models now unified at 0.15/call,regardlessofdurationororientation!HDmodelsora−2−pro(0.15/call, regardless of duration or orientation! HD model sora-2-pro (0.15/call,regardlessofdurationororientation!HDmodelsora−2−pro(0.8/call) is available via Async API only.

​

Supported Models

​

Video Specifications

• Duration: 10 seconds or 15 seconds (selected via model name)
• Quality: 720P / HD
• Orientation: Vertical or horizontal (selected via model name)
• Watermark: No watermark
• Audio: Supports audio-video synchronization

​

API Method Selection

Recommended to use Async API, especially in production environments or when batch generating videos, for better stability.

​

Getting Started

​

Sync API

Quick start with streaming output (for testing)

​

Async API

Task queue, more stable (recommended for production)

​

Character Creation

Extract characters, reuse via @username

​

Model Pricing

Learn detailed model comparison and pricing

​

Code Examples

View multi-language example code

​

FAQ

Solve common issues

​

Important Notes

**Video storage valid for only 1 day!**Generated videos are stored on domestic CDN for convenient access by domestic users. Please download and save to local storage immediately after generation. Both models have extremely high stability. Choose horizontal or vertical mode based on your video display scenario.