Preview and Render Flow

Fresh

Overview

Many Creatify endpoints support a two-phase generation process: first generate a cheaper preview, then render the final high-quality version. This pattern saves credits and allows review before committing to full render costs.

The Pattern

Endpoints That Support Preview + Render

Endpoint	Preview Cost	Render Cost	Total	Direct Cost
URL to Video	1/30s	4/30s	5/30s	5/30s
AI Editing	1/30s	4/30s	5/30s	5/30s
AI Avatar v2	--	4/30s	5+4/30s	5/30s
Custom Templates	1/30s	4/30s	5/30s	5/30s

When to Use Preview

Use preview when:

Testing a new URL or content for the first time
Iterating on creative direction
Budget-conscious and want to review before full render
Generating multiple variations to pick from

Use direct create when:

You have a proven template/content combination
Speed is more important than cost optimization
The content has been previewed successfully before

Preview List (Async) Pattern

Some endpoints support generating multiple preview variations at once:

Code Example: Preview-Render Pattern

python

import time
import requests

def preview_and_render(endpoint, create_params, headers, poll_interval=10, timeout=600):
    """
    Generic preview-and-render pattern for any Creatify endpoint.

    Args:
        endpoint: API endpoint path (e.g., "url_to_videos")
        create_params: dict of creation parameters
        headers: auth headers
        poll_interval: seconds between polls
        timeout: max seconds to wait
    """
    base = f"https://api.creatify.ai/api/{endpoint}"

    # Step 1: Create task (generates preview)
    task = requests.post(f"{base}/", headers=headers, json=create_params).json()
    task_id = task["id"]
    print(f"Task created: {task_id}")

    # Step 2: Poll for preview completion
    result = poll_until_done(f"{base}/{task_id}/", headers, poll_interval, timeout)
    print(f"Preview ready. Previews: {len(result.get('previews', []))}")

    # Step 3: Review (in production, this would be a human review step)
    previews = result.get("previews", [])
    selected = previews[0]["id"] if previews else None

    # Step 4: Render
    render_params = {"preview_id": selected} if selected else {}
    requests.post(f"{base}/{task_id}/render/", headers=headers, json=render_params)
    print("Render started...")

    # Step 5: Poll for render completion
    final = poll_until_done(f"{base}/{task_id}/", headers, poll_interval, timeout)
    print(f"Final video: {final['output']}")
    return final


def poll_until_done(url, headers, interval=10, timeout=600):
    """Poll a task URL until status is done or failed."""
    elapsed = 0
    while elapsed < timeout:
        result = requests.get(url, headers=headers).json()
        if result["status"] == "done":
            return result
        if result["status"] == "failed":
            raise Exception(f"Failed: {result.get('failed_reason')}")
        time.sleep(interval)
        elapsed += interval
    raise TimeoutError("Polling timeout exceeded")

Error Handling

Stage	Error	Recovery
Preview creation	402 Insufficient credits	Check balance, top up
Preview polling	Timeout	Increase timeout or retry later
Preview polling	`failed` status	Read `failed_reason`, adjust inputs
Render	Invalid preview ID	Re-fetch task to get current preview IDs
Render polling	Timeout	Renders may take longer; increase timeout

Preview and Render Flow ​

Overview ​

The Pattern ​

Endpoints That Support Preview + Render ​

When to Use Preview ​

Preview List (Async) Pattern ​

Code Example: Preview-Render Pattern ​

Error Handling ​