Sub-100ms Scraping Speed and Billing Precision
This release introduces a massive boost to overall scraping performance, reducing latency for basic requests down to sub-100ms and accelerating structured data extraction. We have also refined our billing system to ensure real-time spend limit updates and precision refund retries under any server conditions. Finally, support ticket attachment handling and security have been fully upgraded for a more seamless and secure troubleshooting experience.
We moved structured data conversion off the API response path and directly cached the outputs. This eliminates redundant processing and cuts response latency by up to 8 seconds for cached structured results.
We optimized proxy allocation rules to eliminate a 1-second latency floor for basic scraping tasks. Requests that do not require custom proxies are now completed in just 40 to 109 milliseconds.
Our connection multiplexing for synchronous scrapes has been fully redesigned. This significantly reduces connection bottlenecks and improves query response times during heavy parallel scraping workloads.
Improvements
6Accelerated Structured Extraction
We moved structured data conversion off the API response path and directly cached the outputs. This eliminates redundant processing and cuts response latency by up to 8 seconds for cached structured results.
Sub-100ms Scraping Latency
We optimized proxy allocation rules to eliminate a 1-second latency floor for basic scraping tasks. Requests that do not require custom proxies are now completed in just 40 to 109 milliseconds.
Synchronous Scraping Scale-up
Our connection multiplexing for synchronous scrapes has been fully redesigned. This significantly reduces connection bottlenecks and improves query response times during heavy parallel scraping workloads.
Fast-Path Structured Extraction
Lightweight extraction tasks under 50 KiB are now processed directly within the scraping stage, eliminating unnecessary staging hops. This reduces delivery overhead and ensures faster turnaround for small-scale datasets.
Instant Credential Validation
API credential validation is now optimized to run directly from high-speed cache memory without hitting primary databases on subsequent requests. This lowers authentication latency while maintaining real-time revocation capabilities.
Localized Geotargeted Headers
Scraped assets and headers are now automatically built using the correct country context matching your assigned proxy. This ensures highly localized and accurate web scraping results across all target regions.
Bug Fixes
6Instant Spend Limit Releases
Active spending reservations are now instantly released when an API request fails validation checks. This guarantees that failed or rejected requests never artificially count against your spending or concurrency limits.
Preflight Webhook Safety
Webhook URL validation is now completed before any charges or debits are applied to your account. This ensures invalid destination URLs are rejected immediately with no temporary billing impact.
Automatic Disruption Protection
Active scraping jobs are now automatically and safely requeued if a processing node undergoes scheduled maintenance or shutdown. Your jobs are guaranteed to finish executing seamlessly without being dropped.
Better Ticket Attachment Flows
Support ticket submissions and replies now validate attachments before sending, preventing silent file failures. You will receive real-time warnings, accessible error feedback, and retry controls for any failed uploads.
Accurate Search Fallback Deadlines
Optimized search engine fallback latency by enforcing a strict 25-second deadline for parallel dispatch, fallback, and parsing. This ensures your search API calls reliably return responses within the 30-second client window.
General Improvements and Fixes
This release also includes several general reliability enhancements. These include automatic background retries for failed search refunds, accessibility labeling improvements for support screens, and corrected billing scopes for multi-user accounts.
Security
1Improved Attachment Security
Enforced strict, normalized ownership scopes on support ticket attachments. Unauthorized attempts to query or access ticket media now return a generic 404 error, fully protecting your uploaded support files.
Plus 32 internal changes for stability and performance.