Envizion AI API Documentation
Build AI-powered video production pipelines with the Envizion AI API. Create agents, execute runs, render videos, and integrate with webhooks -- all through a single REST API.
Getting Started
Create your first API key and make your first request in under 2 minutes.
Authentication
Learn about API keys, JWT tokens, scopes, and how to secure your integration.
Endpoints
Complete reference for all API endpoints: agents, runs, tools, marketplace, and more.
SDKs
Official Python and TypeScript SDKs with type-safe clients and streaming support.
Webhooks
Receive real-time notifications when runs complete, renders finish, and more.
Rate Limits
Understand rate limiting tiers, headers, and best practices to avoid throttling.
Quickstart
1.Get an API Key
Create an API key from your dashboard, or use the API to create one programmatically. Your key will start with vk_.
2.Make Your First Request
List all available agents:
curl -s https://api.envizion.ai/v1/agents \
-H "X-API-Key: vk_your_api_key"3.Start an Agent Run
Execute an agent with inputs:
curl -s -X POST https://api.envizion.ai/v1/runs \
-H "X-API-Key: vk_your_api_key" \
-H "Content-Type: application/json" \
-d '{
"agent_id": "creative_director",
"inputs": {
"topic": "SpaceX Starship launch highlights",
"style": "cinematic"
}
}'4.Stream Progress
Subscribe to real-time progress via Server-Sent Events:
curl -N https://api.envizion.ai/v1/runs/{run_id}/stream \
-H "X-API-Key: vk_your_api_key"5.Download the Result
Once the run completes, get a presigned download URL:
curl -s https://api.envizion.ai/v1/downloads/projects/{project_uuid} \
-H "X-API-Key: vk_your_api_key"SDK Quickstart
Python
pip install envizion
from envizion import EnvizionClient
client = EnvizionClient(api_key="vk_...")
agents = client.agents.list()
print(agents)TypeScript
npm install @envizion/sdk
import { EnvizionClient } from "@envizion/sdk";
const client = new EnvizionClient("vk_...");
const agents = await client.agents.list();
console.log(agents);See the full SDK reference for complete documentation.
Key Concepts
Agents
Agents are AI-powered pipelines that combine tools to accomplish tasks. Each agent has a set of tools, a pipeline type (simple, sequential, graph, or DAG), and configuration. You can create custom agents or install pre-built templates from the marketplace.
Runs
A run is a single execution of an agent. When you start a run, the agent processes your inputs through its tool pipeline. You can track progress in real-time via SSE streaming, poll for status, or receive webhook notifications when the run completes.
Tools
Tools are individual capabilities that agents use: script writing, asset sourcing, voiceover generation, video rendering, and 48 more. The platform includes 52 built-in tools covering the full video production pipeline.
Credits
API operations consume credits based on complexity. You can check your balance, view transaction history, and receive low-balance webhook notifications. Credits are tied to your plan tier (Free, Starter, Pro).