The Remote MCP Integration enables AI assistants to call SearchAPI tools directly through the Model Context Protocol."Find round-trip flights from NYC to Tokyo in March, under $1000, with one checked bag"
We're starting with travel (flights, hotels, vacation rentals), but this is just the beginning. We have 100+ APIs and will continue adding tools based on your needs.
What's a bundle?
A bundle is a collection of tools you connect to your AI assistant. We offer preconfigured bundles, such as Google Travel, but you can also build your own. Pick exactly the tools you need and create a custom bundle for your workflow.What we launched:
The Walmart Reviews API returns structured review, sentiment, and media data for a single Walmart product. It includes:
Product Data - The product object provides core fields such as title, product link, average rating, total review count, number of text reviews, recommendation percentage, media count, and a full star-rating histogram with per-rating counts and percentages.
Customer Reviews - The reviews.customer_reviews array returns individual reviews with review ID, title (when available), full text, rating, review dates (human-readable and ISO), user identifiers, user name, feedback counts, verification badges, seller information, fulfillment type, and item-level metadata.
Top Reviews - Dedicated top_positive_review and top_negative_review objects return representative customer feedback, including full review text, rating, timestamps, and reviewer metadata.
Aspect Statistics - The aspect_stats array highlights frequently mentioned product aspects (e.g. battery, handle, scratches) with occurrence counts, enabling topic-based analysis and filtering.
AI Review Summary - The ai_summary object provides an overview, including a title, natural-language summary, and aspect-level sentiment spans with polarity indicators.
Bullet Review Summary - The bullet_review_summary object groups feedback into pros, neutral points, and cons, each with standardized aspect names, polarity, and short summaries.
Review Media - The product_media array returns images submitted with reviews, including normal and thumbnail URLs and associated ratings.
Reviews - The reviews array returns individual page reviews with review ID, review link, recommendation status (recommend/doesn’t recommend), full review text, and review dates in both human-readable and ISO formats.
User Information - Each review includes a nested user object containing user ID, name, profile link (when available), and thumbnail image URL.
Engagement Metrics - Reviews return reactions_count and comments_count, allowing analysis of engagement and visibility per review.
HTTP Post Request
HTTP POST request support was added to address pagination limitations when working with large result sets.
When paginating through extensive datasets, the next_page_token can grow significantly in size (for example, 2KB → 5KB → 8KB+ per page). Passing large pagination tokens as URL query parameters can exceed URL length limits enforced by HTTP clients, browsers, proxies, CDNs, load balancers, or servers. This may result in errors such as 414 (URI Too Long) or 413 (Request Entity Too Large).
To mitigate this, POST requests allow pagination tokens to be sent in the JSON request body rather than the URL. This avoids URL size constraints and enables reliable pagination regardless of token length.
This approach is especially important for APIs with opaque or encoded pagination tokens (such as the Meta Ads Library API), where token size is not predictable and can increase across pages.
By supporting HTTP POST, pagination remains stable and consistent even when iterating through very large datasets.
🔧 Engine Improvements:
Google Search API
Fixed ai_overview selectors and improved extraction reliability
Fixed knowledge_graph selectors
Fixed related_questions selectors
Fixed inline_images selectors
Google AI Mode API
Fixed text_blocks.reference_indexes selectors
Fixed q request parameter encoding
Google Images Light API
Fixed an edge case that returned empty results
Google Lens API
Added has_exact_matches key when search_type=all
Fixed issues with incorrectly encoded URLs
Fixed country request parameter handling
Google Shopping API
Fixed shopping_results.product_id and shopping_results.prds values
Google Product API
Fixed an issue returning empty results
Google Product Offers API
Fixed offers.extracted_delivery_price selectors
Fixed offers.delivery_price selectors
Added new keys: promo_code, promo_discount, extracted_promo_discount, promo_expires
Google Maps API
Fixed edge cases when results were empty
Fixed ads selectors
Google Maps Reviews API
Fixed topics extraction
Google News API
Added new countries for hl request parameter validation
Bing Search API
Fixed local_results selectors
Fixed organic_results selectors
Fixed knowledge_graph selectors
Bing Images API
Fixed images.source.link selectors
Airbnb Search API
Improved edge cases when results were not returned
Apple Product API
Fixed app_details.logo cases
Fixed product.videos extraction
Amazon Product API
Fixed product.variants cases
Amazon Offers API
Fixed an edge case returning empty results
eBay Search API
Fixed organic_results.reviews and organic_results.price extraction
eBay Product API
Added a fallback for product.product_id retrieval
YouTube Transcripts API
Fixed requests returning empty results
TikTok Ads Library API
Fixed an edge case where responses returned no results
