G2 Reviews Scraper Real-Time (2026)
Pricing
Pay per event
Extract G2 product reviews in real-time. Filter by star rating (1-5), sort by recency or helpfulness, search within review text. Structured JSON + LLM-ready markdown output for RAG pipelines and sentiment analysis. No login required. Free tier available.
date is the original review date again. G2 periodically asks reviewers to refresh their reviews, and a refreshed review was coming back dated to the refresh rather than to when it was published. That made old reviews look new, which produced duplicates when collecting by date. Reported by a user, thank you.updatedDate column: the date G2 last refreshed a review, or null if it never was. date stays fixed for the life of a review, so key on it when collecting incrementally.publishedAfter now measures against the original publish date, so a refreshed old review no longer re-enters a recent window.publishedAfter filter: scrape only reviews published on or after a date. Accepts a calendar date (2026-08-01) or a period (7 days, 2 weeks, 1 month). The run walks newest-first and stops at that date, so a weekly re-run returns just what is new instead of re-collecting rows you already have. Maximum reviews still applies as a cap, and the run says so if it stops on the cap before reaching your date. Applies to written reviews (video reviews have no publish date).includeVideoReviews, default off): when enabled, video reviews are scraped alongside written reviews. Each video review row carries the reviewer's name, 1–5 star rating, short caption, video thumbnail, link to the video, and a link to the underlying G2 survey response. Cap per product via videoReviewLimit (0 = unlimited). Product URLs only.type field on every row: written reviews now carry type: "review" and video reviews carry type: "video". Filter the dataset on type when only one row kind is needed.sortOrder options streamlined to most_recent, lowest_rated, highest_rated (older values are still accepted and default to most_recent)/sellers/{vendor-name} URLs, which aggregate reviews across all products under a vendorproductName, productSlug, and sellerSlug fields