```yaml
product: AlterLab
title: Caching
category: guides
comparison_context: "AlterLab is an alternative to Firecrawl, ScrapingBee, and Bright Data."
source_url: https://alterlab.io/docs/guides/caching
```

# Caching

Reduce costs and improve performance by caching scrape results. Pay once, reuse many times.

> Caching is **disabled by default**. You must explicitly enable it with `cache: true`.

## How Caching Works

1. **First Request** -- With `cache: true`, the page is scraped, result is stored, and you're charged normally
2. **Subsequent Requests** -- Within the TTL window, same URL requests return cached data instantly. Cache hits are **free**
3. **Cache Expiry** -- After TTL expires, the next request fetches fresh data and refreshes the cache

## Enabling Cache

```bash
curl -X POST https://api.alterlab.io/api/v1/scrape \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/products",
    "cache": true
  }'
```

The response includes `"cached": true/false` and `"credits_used": 0` for cache hits.

## Cache TTL

Set a custom time-to-live in seconds:

```json
{
  "url": "https://example.com/products",
  "cache": true,
  "cache_ttl": 3600
}
```

Default TTL is 15 minutes (900 seconds).

## Force Refresh

Bypass cache and fetch fresh content:

```json
{
  "url": "https://example.com/products",
  "cache": true,
  "cache_force_refresh": true
}
```

This fetches fresh content and updates the cached version.

## Cache Response Headers

| Header | Values | Description |
|--------|--------|-------------|
| `X-Cache` | `HIT` / `MISS` | Whether result came from cache |
| `X-Cache-TTL` | seconds | Remaining TTL for cached result |

## Best Practices

1. **Use caching for monitoring** -- If you scrape the same pages repeatedly, caching can reduce costs by 90%+
2. **Set appropriate TTLs** -- Use shorter TTLs for frequently changing pages, longer for static content
3. **Force refresh strategically** -- Use when you know content has changed
4. **Combine with webhooks** -- Cache results from async jobs for quick retrieval