vcursor-cli 1.0.20

VCursor SDK & CLI - Generate videos from text, images, and URLs using AI

VCursor Python SDK & CLI

Generate videos from text, images, and URLs using AI.

Installation

pip install vcursor-cli

Quick Start

As a library (embedded in code)

from vcursor import VCursor

client = VCursor(api_key="vk-...")

# Submit a video generation task
resp = client.submit("a cat playing piano in a jazz club")

# Wait for completion with progress callback
result = client.wait_for_completion(
    resp.task_id,
    on_progress=lambda p: print(f"{p.data.progress}%")
)

print(result.data.products.url)

Async usage

import asyncio
from vcursor import VCursorAsync

async def main():
    async with VCursorAsync(api_key="vk-...") as client:
        resp = await client.submit("sunset timelapse over mountains")
        result = await client.wait_for_completion(resp.task_id)
        print(result.data.products.url)

asyncio.run(main())

Concurrency limiting

The server enforces per-user rate limits based on subscription tier:

Tier	Standard Requests	Agent Requests	Window
Free	3	1	Lifetime
Pro	30/hr	5/hr	Rolling 1h
Plus	120/hr	15/hr	Rolling 1h

# Check current rate limit status
status = client.check_concurrency("standard")
print(f"Standard: {status.summary['standard'].used}/{status.summary['standard'].limit}")
print(f"Agent:    {status.summary['agent'].used}/{status.summary['agent'].limit}")

# Client-side limit (capped at server limit)
client = VCursor(api_key="vk-...", max_concurrency=10)

As a CLI

# Authenticate
vcursor login

# Generate a video
vcursor generate "a cat playing piano"

# Agent mode
vcursor generate --agent "create a 30s commercial"

# Check concurrency status
vcursor concurrency

# With concurrency limit
vcursor generate --max-concurrency 2 "sunset timelapse"

API Reference

VCursor / VCursorAsync

Method	Description
submit(prompt, **kw)	Submit a standard video generation task
get_progress(task_id)	Get task progress
wait_for_completion()	Poll until task completes
cancel_task(task_id)	Cancel a running task
submit_agent(message)	Submit an agent mode task
get_agent_progress()	Get agent task progress
wait_for_agent_completion()	Poll until agent task completes
check_concurrency()	Check concurrency limit status

Configuration

Config is stored in ~/.vcursor/config.json (shared with the Node.js CLI).

Environment variables:

• VCURSOR_API_KEY - API key
• VCURSOR_SERVER - Server URL

License

MIT