Async Video Generation

Video generation can take several minutes. Synchronous requests risk 502 timeouts from proxies, CDNs, or serverless function limits. Async mode solves this.

The Problem

When using a server proxy (Cloudflare, Vercel, etc.), requests that take longer than the proxy timeout (typically 30-60 seconds) will be killed with a 502 error. Video generation routinely exceeds this.

The Solution: Async Mode

from visgate_sdk import AsyncClient
import asyncio
 
async def generate_video():
    async with AsyncClient() as client:
        result = await client.videos.generate(
            model="fal-ai/veo3",
            prompt="a bird flying over mountains",
            duration_seconds=6.0,
        )
 
        # Check if the video is ready
        if result.video_url:
            print(f"Video ready: {result.video_url}")
        else:
            print("Video is still processing...")
 
asyncio.run(generate_video())

Request Status Properties

Property	Type	Description
`requestId`	`string`	Unique request identifier
`status`	`string`	`"pending"` \| `"processing"` \| `"completed"` \| `"failed"`
`mediaType`	`string`	`"image"` or `"video"`
`provider`	`string`	Provider name
`model`	`string`	Model identifier
`outputUrl`	`string \| null`	Output URL when completed
`errorMessage`	`string \| null`	Error message if failed
`createdAt`	`string`	ISO timestamp of creation
`completedAt`	`string \| null`	ISO timestamp of completion

Webhooks (Alternative)

Instead of polling, you can provide a webhook URL to get notified when the job completes:

const result = await client.videos.generate("fal-ai/veo3", "ocean waves", {
  preferAsync: true,
  webhookUrl: "https://your-app.com/api/webhook",
  callbackId: "my-custom-id",
});

The webhook will receive a POST request with the completed result and your callbackId.

Proxy Configuration

Make sure your server proxy route allows enough time:

app/api/visgate/[...path]/route.ts

export const maxDuration = 300; // 5 minutes
export { GET, POST, PUT, DELETE, OPTIONS } from "@visgate_ai/server-proxy/next";

The Problem

The Solution: Async Mode

Request Status Properties

Webhooks (Alternative)

Proxy Configuration

On this page