: Support for running custom prompts against crawl data using OpenAI and Google Gemini (expanded in versions 21 and 22).
This minor update targeted a very specific problem affecting users who relied on automated reporting workflows to Google's Looker Studio. By resolving this issue, version 19.8 delivered a more reliable and predictable experience for professionals who depend on scheduled exports for client dashboards.
Better detection of content hidden behind user interactions. Reduced memory overhead during heavy JS crawls. Integration and API Connectivity Screaming Frog SEO Spider 19.8
| Limitation | Workaround | | :--- | :--- | | due to DNS resolution overhead | Use "Subdomain Discovery" sparingly. Pre-seed a subdomain list via a text file. | | JavaScript rendering max concurrent threads capped at 8 to avoid Chrome instability | Use "Sequential JS Rendering" mode for highly unstable sites. | | No native ARM64 build for Mac (runs under Rosetta 2) | Acceptable performance loss (~10-15%). Use Windows VM for ARM Windows builds. | | Cannot parse log files larger than 50GB on 16GB RAM machines | Split logs using command-line split tool before loading. | | Database mode (SQLite) not available in Free version | Upgrade to Paid or limit crawl depth to 3. |
:
Spider > Rendering:
The power of Screaming Frog lies in its ability to extract custom data via XPath, CSS Path, and Regex. Version 19.8 fine-tunes the parsing logic for these extraction methods. : Support for running custom prompts against crawl
: To pull user behavior and traffic data directly into your crawl reports.
Spider > Crawl Limits:
: