How to Slash Your Gaffa Credit Costs by 40+%%
Learn practical tips and settings to cut your Gaffa credit usage, so you get more out of every request without sacrificing results.
Mar 19, 2026
Modern websites are packed with high-resolution images, videos, and interactive media. This results in engaging sites, but when you're scraping for text or data, every image, video thumbnail, and banner ad consumes bandwidth, slows your requests, and costs you credits even though you don't actually need them.
Using Gaffa's Browser Request API, each request starts with a fresh browser context, so there's no cached data. Every image is downloaded from scratch, every time. If you're scraping 1,000 news articles and each article page downloads 5 MB of images you'll never use, that's 5GB of wasted bandwidth costing you credits.
To solve this, Gaffa's max_media_bandwidth setting lets you cap or completely block image and video downloads, dramatically reducing costs while still allowing you to scrape all the text data you need.
What is max_media_bandwidth?
The max_media_bandwidth parameter controls the maximum amount of media content (images and videos) your browser downloads. The settings have three states:
- null (default): No limit (all images and videos will be downloaded)
- 0: Block all images and videos completely
- Any whole number (e.g., 5): Cap media downloads at that many megabytes
Here's how you use it in a Gaffa request:
sample-request-with-max_media_bandwidth
When `max_media_bandwidth` is set to 0, Gaffa monitors all downloads and cancels any image or video requests, whilst still allowing HTML, CSS, JavaScript, and other essential page resources to load.
Testing max_media_bandwidth on BBC.com
To demonstrate the real-world impact of this setting, we ran four separate tests on BBC.com, one of the world's most popular news sites. The BBC homepage features dozens of article thumbnails, video previews, and banner images, making it an ideal candidate for testing image-blocking optimization.
For each test, we:
- Captured a screenshot to show the visual difference
- Generated markdown to extract the text content
- Measured credits used and request time
Methodology
We ran the same request eight times on BBC.com: four with images enabled (without max_media_bandwidth) and four with images blocked (with max_media_bandwidth set to 0):
Scraping BBC.com with max_media_bandwidth
Results
Test URL: https://www.bbc.com/
Screenshot - WITH Images (when max_media_bandwidth is null)
BBC screenshot from Gaffa when max_media_bandwidth is null
Screenshot - WITHOUT Images (when max_media_bandwidth is 0)
BBC screenshot from Gaffa where max_media_bandwidth is 0
Individual Test Results
Test 1
Credits Used
Without max media
With max media
Savings
14 credits
8 credits
6 credits
Request Time
Without max media
With max media
Savings
5 seconds
3 seconds
2 seconds
Test 2
Credits Used
Without max media
With max media
Savings
11 credits
7 credits
4 credits
Request Time
Without max media
With max media
Savings
4.8 seconds
2.9 seconds
1.9 seconds
Test 3
Credits Used
Without max media
With max media
Savings
12 credits
7 credits
5 credits
Request Time
Without max media
With max media
Savings
5.2 seconds
3.1 seconds
2.1 seconds
Test 4
Credits Used
Without max media
With max media
Savings
12 credits
6 credits
6 credits
Request Time
Without max media
With max media
Savings
4.9 seconds
3 seconds
1.9 seconds
Average Results Across All Tests
Credits Used
12.25 credits
With max_media_bandwidth: 0
Average Savings
7 credits
5.25 credits (43%)
Request Time
5 seconds
With max_media_bandwidth: 0
Average Savings
3 seconds
2 seconds (40%)
Key Finding: By blocking images on BBC.com, we achieved an average credit reduction of 43%, requests completed 40% faster, and 100% of the text content was captured, all without impacting the article headlines, text, or metadata we were trying to extract.
When should you use max_media_bandwidth?
1. News Article Scraping: When monitoring news sites or building news aggregators, you only need article text, headlines, and publication dates, not the accompanying photos.
2. E-commerce Price Tracking: Track product prices and descriptions across multiple sites. You don't need product images, just the text data.
3. Review Mining: Extracting customer reviews from e-commerce sites, travel sites, or app stores, the text reviews are what matter, not user profile pictures.
4. SEO Analysis: When analyzing page titles, meta descriptions, headings, and content structure, images are irrelevant.
5. Bulk Content Extraction: Scraping hundreds or thousands of blog posts, articles, or forum threads where only text content matters.
When not to use max_media_bandwidth
Don't use max_media_bandwidth when:
- You need to capture screenshots for visual verification
- You're testing page layouts or responsive design
- You need to verify that specific images are present
- You're analyzing image content itself (alt text, file sizes, etc.)
- The site uses images to load critical content (rare, but possible)
Best Practices
Based on our testing and real-world usage, here are our recommendations:
1. Start with Zero, Adjust If Needed
Begin with max_media_bandwidth: 0, and increase only if you encounter issues. Most text extraction works well without images.
2. Monitor Your Savings
Track your credit usage before and after implementing max_media_bandwidth to quantify ROI. Most users see a 40-50% reduction in costs on image-heavy sites.
For most web scraping and automation tasks, news articles, product descriptions, reviews, or any structured text content, images are unnecessary overhead. Setting max_media_bandwidth to 0 is the simplest optimisation you can make: start with it enabled, test that your data quality remains unchanged, and monitor the difference in your credit usage. Based on our testing, most users working with image-heavy sites will see savings of 40–50%. Try it in the API Playground today.
Ready to build your own integration?
Join hundreds of teams using Gaffa to automate their browser workflows