Releasesby GlyphCustom browser instructions for LLM sources
LLM sources now support custom browser instructions, allowing you to define actions (e.g., clicks, scrolls, waits) to control page behavior before content is extracted.
Also shipped Jul 24, 2026Oxylabs in July 2026 →
Skill and plugin sharing in CoCo (Public Preview)
Skill and plugin sharing is now available in public preview. Share, discover, and govern CoCo skills and plugins across your Snowflake account. Publish them as Cortex Extension objects that teammates can find and install, with administrator certification, access control, and usage telemetry. For more information, see Share skills and plugins .
Governed tags system table is in Beta
The system.tags.governed_tags system table is in Beta . This table contains a record for each governed tag key in your account, including customer-created and system-generated tags. Use it to programmatically determine which tags in your account are governed. See Governed tags system table reference .
Git CLI commands in Git folders are now in Public Preview
Git CLI commands in Git folders are now in Public Preview . Git folders that you create from the UI automatically get Git CLI access when your workspace is eligible, so you can run standard Git commands such as git stash , interactive rebase, pre-commit hooks, Git submodules, and Large File Storage (LFS) directly from a Databricks terminal.
Diagnose and fix Lakebase issues with Genie, using telemetry in Unity Catalog (Beta)
Lakebase adds AI-assisted troubleshooting: database issues and early warnings are surfaced to you with Insights , and you can then investigate and fix them conversationally with Genie . Both draw on database telemetry captured to Unity Catalog. See Observability .
Knowledge Catalog governance workflows for data product access in preview
Governance workflows add automated controls for data product access management through a request-review mechanism.
Claude Opus 5 launches with a 1M token context window at $5 / $25 per MTok
Claude Opus 5 improves on Opus 4.8 with a 1M token context window, 128k max output and thinking on by default, at the same $5 / $25 per MTok price. It is available on the Claude API, Bedrock, Google Cloud and Microsoft Foundry.
More from OxylabsEvery Oxylabs release →
Custom parser for Home Depot data
A dedicated parser for Home Depot product pages is now available through the universal scraping source. Enable parsing on your request to receive structured product data (such as title, price, availability, and specifications) instead of raw HTML.
Structured Google Scholar results
Google Scholar is now a dedicated parsed source in Web Scraper API. Set source to google_scholar and parse to true to receive parsed JSON directly.
Increased prompt limit for Google AI Mode
The maximum prompt length for Google AI Mode source requests has increased from 4,000 to 8,000 characters.
Improved success rates on hard-to-reach targets
We have rolled out improvements to how requests are handled on sites using CAPTCHA. Success rates on these targets should improve, particularly for jobs that previously saw elevated failure or retry rates. Learn more in our documentation.
More accurate Walmart e-commerce pricing
Parsed output from Walmart e-commerce sources now returns pricing data more reliably and consistently. If you're tracking competitor prices or monitoring product catalogs, you'll get cleaner numbers automatically with new requests using "parse": true.
Improved fingerprint validation and device-memory alignment
Headless Browser now offers more realistic browser fingerprints and continues to align device-memory values with the rest of the emulated fingerprint to lower detection risk for all users. This change applies automatically to all Headless Browser sessions.
Oxylabs Agent Skills
Oxylabs now offers official Agent Skills – structured skill files that teach AI agents and coding assistants work with Oxylabs products with significantly more accurate code and fewer hallucinations.
Citation retries for ChatGPT web search
ChatGPT source requests with search set to true now retry citation extraction automatically, roughly doubling citation coverage in the parsed citations field. The overall request success rate is unchanged, and no payload changes are needed for requests already using search: true.