Scrapeless Weekly Changelog: September 7–13, 2026

TL;DR:

• Google Search API documentation is live. Developers now have a dedicated introduction and request reference for structured Google Search data.
• ChatGPT Scraper adds request-level web search control. The web_search parameter lets each task choose whether to use web search.
• ChatGPT responses now include conversation_id. The new field makes it easier to associate stored results with their source conversation.
• ChatGPT Scraper adds HTML and SSE output options. Teams can select the response representation that fits their processing workflow.
• Language and task-routing issues were fixed. Updates cover Perplexity fanout data and intermittent “plan not found” errors across AI Scraper.

---

The September 7–13 release gives developers more control over AI-answer collection and a clearer path into Google Search data. New Google Search API documentation is now available, and ChatGPT Scraper gains web search control, conversation identifiers, and two additional output options. The release also corrects language-selection and task-routing issues across AI Scraper.

🌟 Google Search API Documentation Is Live

The new Google Search API documentation introduces the dedicated API surface for extracting structured Google Search data in real time.

The documentation covers common request controls such as query, country, language, Google domain, result type, pagination, and result count. It gives teams a single reference for designing search-monitoring, research, and data-enrichment workflows without maintaining their own browser and parsing stack.

Google Search API remains available through Scrapeless’s search-data products. Teams planning a new integration can review the Google Search API product page alongside the newly published reference.

💫 ChatGPT Scraper Adds More Request and Output Control

This week’s ChatGPT Scraper update covers how a task uses web search, how its conversation is identified, and how response data can be delivered.

Choose Web Search Per Request

The scraper.chatgpt actor now supports custom web_search input. Set the boolean for each request to choose whether ChatGPT web search is enabled for that task. This is useful when a dataset mixes model-only prompts with prompts that need current web information and source results.

The ChatGPT Scraper reference documents web_search as an optional boolean. The wider behavior of web-connected answers and citations is described in the official web search guide.

Identify the Source Conversation

ChatGPT Scraper responses now include conversation_id. Data teams can store the identifier with the answer, prompt, citations, and collection timestamp so records from the same source conversation are easier to trace and group.

The field identifies the conversation associated with the response. This release note does not assume that it can be passed back to continue a conversation; integrations should follow the current request documentation for supported inputs.

Select HTML or SSE Output

ChatGPT Scraper can now provide HTML and SSE output formats. HTML is useful when an application needs the rendered response representation. SSE is suited to workflows that consume the underlying event stream or preserve it for later processing.

The HTML Living Standard defines the document syntax applications can use when validating or transforming returned HTML. SSE consumers should parse the stream as event data rather than treating it as one HTML document. The Server-Sent Events specification defines the event-stream model and parsing rules used by standards-based clients.

For a broader overview of structured ChatGPT answers and citations, see the ChatGPT collection guide.

Start Scraping with Scrapeless

Power up your web scraping and automation workflow with Scrapeless!
Sign up today and get $5 in free credit — no credit card required.

Claim your free credit now in the Scrapeless Dashboard.

⬆️ AI Scraper Reliability Fixes

Two fixes improve consistency across AI Scraper responses:

• Perplexity fanout language handling. Fanout data no longer returns the incorrect language in the affected cases.
• Intermittent task-routing errors. The occasional “plan not found” response affecting AI Scraper tasks has been fixed.

These are service-side fixes. Existing integrations do not need a request-schema migration for the corrected behavior.

📍Conclusion

This release makes Google Search API easier to adopt and expands the ways teams can collect and manage AI answers. ChatGPT Scraper now offers per-request web search control, conversation identifiers, and HTML or SSE output. The accompanying fixes improve language consistency and task execution across AI Scraper.

AI Scraper is part of the Universal Scraping API. Review Scrapeless pricing when planning production workloads.

👀 Ready to Build Your AI Search Data Workflow?

Join developers building search-data and AI-answer workflows in the Scrapeless community: Discord · Telegram.

Sign up at app.scrapeless.com to get started.