Logimu

ACCOUNT: the account's balance, plan, 30-day usage by endpoint and its product-tracking allowance (snapshots a day: total, used, remaining) - the same as REST GET /v1/account. USE WHEN the user asks how many credits are left, what a job will leave, what the key has been used for, or how much tracking room there is. COST free, never billed. API key required (keyed lanes); without one, nothing to report.

BATCHES: run any Logimu endpoint over a whole list in one job and come back for the results - the same batch API as REST /v1/batches (see https://api.logimu.com/docs#batch). action=estimate prices a list before anything is billed (free); action=create makes a named, re-runnable batch (add start=true to launch its first run at once); action=run launches a run of an existing batch; action=status reports a batch (runs, line stats) or, with run_id, one run's progress (rows done, credits so far, queue position); action=results returns completed result rows of a run, mid-run too (limit / offset to page; jsonl_url and csv_url give the full download); action=list lists the account's batches; action=schedule repeats a batch daily / weekly / monthly (credit_ceiling, webhook_url); action=delete removes a batch. USE WHEN the user has more than a handful of ASINs, barcodes, search terms, seller ids, brands or seeds to run, wants a daily/weekly refresh, or wants a CSV. Kinds: product (asins or gtins - or source=tracked for the ASINs this account tracks, typically with include [history_only] at 0.05 a row), search (terms, live SERP), seller / brand / category / keywords (paid plans), and every other endpoint as its own kind (shopping, warehouse_search, bestsellers, offers, deals, seller_products, seller_brands, seller_network, brand_sellers, brand_network, keywords_compare, categories, category_search, brands_top, brands_movers, sellers_top, sellers_movers, categories_top) with the endpoint's query parameters in params. Each row bills what the endpoint bills; failed, not-found and duplicate rows are never billed; runs are serialized per marketplace. COST estimate / status / results / list are free; create + run queue billable rows (the estimate is the ceiling). API key required (keyed lanes).

LIVE Amazon Best Sellers list for one category: up to 100 products in Amazon's rank order, the category's full breadcrumb name, and its subcategories with their URLs. Given a marketplace's root best-sellers URL (or just a country) it returns the top-level categories instead, so you can walk the tree: root -> category -> subcategory, always by the URLs a previous answer returned. USE WHEN the user asks what sells best in a category or department right now, wants a top-100 or best-seller list, or wants to browse Amazon's best-seller categories. DON'T USE for a keyword search (serp is Amazon's live results page, shopping a curated shortlist), for one known ASIN (product), or for a product's rank history (product with bsr_history). RETURNS page_type 'category': current_category (name, full_name = the breadcrumb path, path[] with each ancestor's name/url/node_id, node_id, url), subcategories[] (name, url, node_id), bestsellers[] (rank, asin, title, link, image, rating, ratings_total, price) and pagination; page_type 'root': categories[] (name, url, node_id). Every answer carries observed_at (when the list was fetched from Amazon) and cached. HONESTY: names are in the page's language (German pages are local; fr, it, es, jp, mx and br pages are requested in English); ranks are Amazon's at fetch time; price is null when the card shows none; node_id is null for departments whose URL carries only a slug. LATENCY is SECONDS: ~15-35s for a fresh top 100, ~15s for a root list; Amazon refreshes these lists hourly, so an answer may come from our cache (up to 30 min for a category, 24 h for the root list) and then returns in milliseconds. MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br (the marketplace comes from the URL; no Walmart). COST API key required (the free no-signup lane cannot run live fetches). 1 credit per page of 50 products returned (the full top 100 = 2), cached or fresh; the root list is free; a category Amazon does not have, a blocked fetch and an empty list are never billed.

Amazon BRAND dossier from Logimu's own buy-box observations (derived intelligence, not estimates): catalog size, how many sellers win the buy box, control score (share of buy-box days won by FBA offers), winner diversity, Amazon's own retail share, FBA penetration, seller concentration (HHI) and churn, seller-pressure trend grades, since when we have observed the brand, its dominant category with p25/p50/p75 benchmarks, and the top-10 sellers by observed buy-box days with their feedback records. USE WHEN the user asks who sells a brand, how contested / Amazon-dominated / FBA-heavy a brand is, whether a brand is gaining or losing sellers, or wants to vet a brand for wholesale / arbitrage / brand protection. OPTIONS: sellers=true adds the full seller list (25 per page, page=N); network=true adds the brand's SELLER MAP - the brands connected to it because the same third-party sellers win the buy box on both (edges weighted by shared sellers, one hop, max 50); history=true adds up to 24 calendar months (ASINs, sellers, buy-box days, price, rating). All 11 Amazon marketplaces, no Walmart. HONESTY: 'observed since' is our data horizon, not the brand's age; a 404 means we have not observed the brand (names are matched as Amazon prints them). COST: paid plan required (the tool is listed only for keys on a paid plan). 1 credit for the dossier, +0.5 once when history comes back non-empty, +1 per page of sellers, +1 for the network. Unknown brand: 0.

Amazon CATEGORY landscape from Logimu's observations: size (ASINs, brands, sellers), pricing and rating averages, Amazon's own retail share, FBA penetration, Amazon private-label share, seller concentration (HHI top-5), heat (competitive intensity), 30-day catalog churn, demand signals (bought-past-month, revenue signal, velocity tier), root-level benchmark percentiles, the sub-categories and - for root categories - the top-10 brands by observed buy-box days. USE WHEN the user asks how competitive / Amazon-dominated / FBA-heavy a category is, how big it is, which brands lead it, or wants to size a private-label or wholesale entry. Give category_id (Logimu's category id - the one every search result carries; or node_id for an Amazon browse-node id) OR q (a name fragment, English alias included) - with q the tool first returns matching categories (free) so you can pick the id - OR list=true to walk the taxonomy (the marketplace's root categories, or the children of parent_id; free). history=true adds months of products / brands / sellers. All 11 Amazon marketplaces. COST: paid plan required (listed only for keys on a paid plan). 1 credit per category profile, +0.5 with non-empty history; the name lookup and the taxonomy listing are free; unknown id: 0.

LIVE Amazon Today's Deals: the deals page of one marketplace, read from Amazon at request time and filtered the way Amazon's own page filters it - department (category_id), deal collection (deal_types: lightning_deal, trending, coupons, started_today, ending_today; or any collection id an answer lists), percent off, price, Prime Exclusive, minimum rating, brand. USE WHEN the user asks what is on sale or discounted on Amazon right now, wants today's deals or lightning deals in a department, deals above a discount or under a price, coupon deals, or deals ending today. DON'T USE for one product's price history or past price drops (product with offer_history), for who sells a product right now (offers), for a keyword search (serp, shopping) or for what sells best in a category (bestsellers). RETURNS deals[] (position, asin, title, link, image, deal_id, deal_link, deal_type e.g. BEST_DEAL / LIGHTNING_DEAL, is_lightning_deal, deal_state, percent_claimed, ends_at, deal_price and list_price {value, currency, symbol, label}, percent_off, savings, deal_badge, is_prime_price, coupon {text, amount, promotion_id, link}, brand, brand_id, department_ids, product_group, product_type, variations), categories[] (category_id, name, selected: the departments to pass back as category_id), collections[] (id, name: pass back as collection), facets (brands with ids, reviewRating, price and percentOff ranges, accessType), selected_filters, filter_check (per requested filter: applied / not_applied / unverified), pagination (page, page_size, pages_fetched, start_index, returned, has_more, partial), amazon (marketplace_id, locale, currency), observed_at and cached. HONESTY: an empty deals list is a real answer (e.g. no lightning deals running at this hour); a filter Amazon did not apply (filter_check not_applied, e.g. a category_id that is not one of the marketplace's deals departments) is refused as an error, never billed, instead of serving the unfiltered page; unverified filters are served, so say so before relying on them; there is no total count (has_more says whether the last page came back full) and no deal start time; ends_at only where Amazon shows a countdown; list_price is Amazon's reference price with its own label (RRP, Was, Typical ...), null when the page shows none. LATENCY is SECONDS: ~10-30s a page (pages=5 up to ~2 min); an identical request within 15 min may come from our cache in milliseconds. MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br (no Walmart). COST API key required (the free no-signup lane cannot run live fetches). 1 credit per deals page fetched that returned at least one deal, whatever its size (30, 60, 90 or 120 deals are all 1 credit, so page_size=120 gives the most deals per credit), cached or fresh; a page with no deals, an empty list, a blocked or busy fetch and bad input are never billed.

Page 1 of Amazon US search for a keyword, as Logimu last OBSERVED it (head search terms taken from Amazon's own autocomplete, each observed about once a week, all departments, one US delivery ZIP): the organic order with product facts (title, brand, price, rating) and each product's latest observed buy box (seller, AMZ / FBA / FBM, other sellers), the sponsored units present in the page HTML (Sponsored Brands banners and videos, server-rendered Sponsored Products; ones Amazon loads later by JavaScript are counted as slots only), the Overall Pick (Amazon's Choice) holder, Amazon's result count, page_summary (organic slots by brand and by buy-box seller, Amazon's own buy boxes, FBA / FBM, Sponsored Brands units by brand) and the products that entered or left page 1 since the previous observation, all dated. history=true adds every weekly observation and the rank series of the page's products. USE WHEN the user asks which products, brands or sellers hold page 1 for a search term, who advertises on it, or how the page changed. Amazon US only. NOT volume figures (Amazon publishes none for search terms and none is implied) and NOT live: a dated weekly sample, and the same search minutes apart differs by a few products - for a live page use the serp tool, for what Amazon suggests for a seed the keywords tool, for the keywords a product shows up for product_keywords. A keyword outside the observed set is free (keyword_not_found). COST: paid plan required (listed only for keys on a paid plan); 1 credit (+0.5 with history when more than one observation exists). Counts against the derived daily cap (default 2,000 calls per account).

Amazon's OWN autocomplete suggestion lists for a seed term, from Logimu's continuous harvest (since Aug 2026, 11 marketplaces): up to 6 lists - one per crawled prefix / department - each term with its position 1-10 WITHIN that list, when it was first and last observed and the marketplaces it was observed in, plus related observed vocabulary starting with the seed. history=true adds every observation of each list since the harvest began (a term entering, leaving or moving position). compare='us,uk,de' instead returns the seed's primary list on each marketplace and a matrix of positions per marketplace. USE WHEN the user asks what buyers type for X, wants listing / backend keyword language, seeds for product research, or how a term's suggestions differ by country. NOT volume figures (Amazon publishes none for its suggestions and none are implied), NOT a product-ranking map - which products appear on page 1 for a term is keyword_rankings, the keywords a product shows up for is product_keywords. Positions are comparable only within one list. COST: paid plan required (listed only for keys on a paid plan), 1 credit per lookup or compare, +0.5 with history when at least one closed observation exists; an unknown seed is free. Counts against the derived daily cap (default 2,000 calls per account).

LIST the products this API key tracks (track_product): each with its status, interval, expiry, whether the hot lane has picked it up, the lane's last and next read, plus the plan's tracking allowance (snapshots a day: total, used, remaining). USE WHEN the user asks what is being tracked, whether a tracking is live, how much allowance is left, or wants to stop one (the id goes to DELETE /v1/track/{id} on the REST API). Stopped trackings stay listed for 30 days. COST 0 credits. Paid plans only.

Ranked LISTS from Logimu's observations, 25 per page: kind=brand_movers (brands gaining or losing distinct buy-box sellers fastest over 30 or 90 days), seller_movers (sellers whose buy-box footprint grew or shrank most), top_brands (brands ranked by observed buy-box days / sellers / control_score / asins / revenue / amazon_share, optionally within a root category_id), top_sellers (third-party sellers ranked by brands / buybox_days / asins / revenue - the cross-brand operators), top_categories (categories ranked by asins / brands / heat / revenue / price / sellers, root categories by default, or the children of parent_id). USE WHEN the user wants to DISCOVER brands, sellers or categories rather than look one up - sourcing leads, brands under seller pressure, the biggest resellers, which categories are hot. Then drill in with the brand / seller / category tools. All 11 Amazon marketplaces. COST: paid plan required (listed only for keys on a paid plan), 1 credit per page of 25 rows returned (an empty page is free). Daily cap per account on the derived family (default 2,000 calls).

LIVE Amazon offers page for one product: every seller offering it right now, read from Amazon's 'All offers' view at request time, every page of it (up to 10). Each offer has the seller and seller_id, price, shipping_cost, condition, buy box, FBA / FBM / Amazon, Prime, ships-from, the delivery promise and the seller's feedback record; the answer adds the buy box and the page's own header (title, image, rating, review count). No product page is fetched. USE WHEN the user asks who sells a product right now, wants every current offer, the cheapest new or used offer or the buy-box winner as of this minute, or needs a fresh seller list before buying, sourcing or repricing. DON'T USE for brand, BSR, listing content, reviews or price / offer history (product), for a keyword search (shopping, serp) or for a brand's or a seller's long-run buy-box record (brand, seller). RETURNS asin, marketplace, observed_at, title, image_url, rating, reviews_count, currency, price (the buy box), buybox_seller, buybox_seller_id, seller_count, no_sellers, pages_fetched, complete (false = Amazon stopped serving the offers pages part-way: the list is the part read, with partial_reason and a note), buybox_condition / buybox_condition_source / buybox_condition_detail (new | used | null = no featured offer; the pinned offer's condition text or the used storefront's name, e.g. Amazon Resale), current_sellers (sellers[] buy box first, then price ascending: seller, seller_id, price, shipping_cost, condition, condition_class, is_buybox, is_fba, is_prime, fulfillment, ships_from, delivery, seller_rating, plus buybox_days_30d and the 30-day buy-box shares from our observations; offer_count, used_offer_count, lowest_new, lowest_used), fulfillment (offer counts by AMZ / FBA / FBM, whether Amazon sells or holds the buy box) and seller_ratings; with gtin also gtin_resolution (source, confidence, barcode_verified). HONESTY: no_sellers=true means Amazon lists no offer for the product right now, an answer confirmed by a second, independent fetch before it is served; a fetch that fails is reported as an error, never as no sellers. LATENCY is SECONDS: typically 4-12s (one page ~4s; a long list walks every page). MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br (no Walmart). COST API key required (the free no-signup lane cannot run live fetches). 0.5 credit when the answer carries at least one offer; a no-offers answer, an ASIN Amazon does not have, a barcode nothing resolves and a failed or busy fetch are never billed.

Full dossier for ONE known product: its current snapshot plus its observed history. USE WHEN the user has a specific ASIN, Walmart item ID, product link, or a product_id returned by shopping or search, and asks about price history, historical prices, price changes, 30-day history, stock history, seller history, buy-box history, historical analysis, 'analyse this product', 'is this a good buy', 'has the price moved/dropped', 'who is selling this', 'is it in stock'. This is the ONLY tool that returns history: shopping and search return current values, so any historical question about a product they listed comes here. DON'T USE to discover products from a keyword (use shopping) or to pull a filtered list (use search). RETURNS current price, BSR, rating, review count, stock, buy-box seller and seller count, plus an observed_at freshness stamp, full price_history and stock_history back to first observation (keyed; the free lane carries the 30-day views), change events tagged with the buy-box seller at each change, the current all-seller offer table with 30-day buy-box days (each seller row carries fulfillment AMZ/FBA/FBM and delivery days), a fulfillment block (buy-box AMZ/FBA/FBM, offer counts, whether Amazon sells and holds the buy box, buy-box delivery days and dispatch latency), the bought-past-month badge (measured aggregate buyer behavior, not an estimate), brand stats, and since 2026-09-27: dimensions (item weight + package size as printed plus parsed metric weight_g / *_cm), deal_history (list price, savings and coupon/promotion text as segments of consecutive observations, with a current view), bought_past_month_history (the badge over time; null = page seen without a badge), rating_history (rating_count + rating per observed day with ratings_gained_30d/90d), and variation_coverage (the variation matrix as last seen with, per child, whether we track it and its latest 30-day price - current state, not history). With offer_history=true every per-seller point also carries that day's shipping cost, fba and prime, and the answer carries buybox_price_history (2026-10-09): the lowest buy-box price of every observed day, oldest first, each point labelled source offers (our offer observations, every Amazon marketplace, from 2026-04-17) or daily_snapshot (the legacy daily snapshot, Amazon US back to Nov 2023) - the deepest price series we hold; a gap between two dates is a stretch we did not observe. Since 2026-10-14 every Amazon answer also carries buybox_price_points_hot: every all-offers-page observation of the last 30 days, newest last, {ts, price, seller, seller_id, fulfillment, offer_count} - sub-daily for ASINs in our hot lane, which refreshes popular and customer-polled ASINs every few hours (null when none; history_only carries it too) - and tracking: the lane's state for the ASIN (hot, interval_min, last_fetched_at) or null. Since 2026-10-15 Amazon answers carry buybox_condition (new | used | null = no featured offer; with buybox_condition_source product_page | offers_page and buybox_condition_detail: the pinned offer's condition text, used_only_buybox, or the used storefront's name - Amazon Resale / Amazon Warehouse win the buy box like any seller and read as Amazon retail before), buybox_state (what the lane's newest all-offers read saw: in_stock | no_buybox | out_of_stock, with buybox_state_observed_at and stock_state_source when it decided stock_state; stock_state, in_stock and out_of_stock always agree), and on LIVE answers variation_status (page_is_sibling | page_is_other | null) + page_asin when Amazon served another ASIN's page for the one asked (cache answers: null); unit_price on LIVE answers = Amazon's per-unit price of price where the page prints one ({symbol, value, currency, raw, per}; null otherwise); with full_offers, offers_complete (false = the offers pages stopped part-way, offers_partial_reason says why). Amazon answers also carry the observed product-page content, the byline (authors[] with role_norm author/narrator/illustrator/editor/translator + author for books, audiobooks, music; 2026-09-29), the book/media format (format, format_norm, and formats[]: every format of the book, each its own ASIN with price and membership subscription_price, plus family_asin; 2026-09-30) block: description (with description_source), feature_bullets, images, breadcrumbs, variations with variation_count and parent_asin, stamped content_observed_at — content_observed_at:null with empty arrays means the content crawl has not captured this ASIN yet, never 'this product has no description/gallery'. For the ~17% of the catalog with no overall rank (media, books, niche items), bsr_leaf and bsr_leaf_category carry the best category rank instead. Every response carries a data_source field naming the marketplace the numbers were observed on (e.g. 'amazon US marketplace — observed listings') — attribute prices to that source when presenting them; they are marketplace listings, not manufacturer or site-wide prices. MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br, walmart. Walmart takes a numeric item ID (or gtin= with the item's UPC) and returns the intelligence blocks plus its listing content: image_url, description, breadcrumbs, highlights (spec name/value pairs), warnings, upc, rating/review_count, was_price, in_stock, and shipping_cost per seller; mode=live reads the walmart.com page now (2 credits, typically 8-15 s, one page: the cheapest other seller is not fetched; location names the store it was priced for) and mode=auto serves our record when observed within max_age_days, else reads live. COST free lane 1 of 30 daily queries, cache only, and returns the snapshot + 30-day views (the full history streams, bsr_history, offer_history and live scrapes need an API key (plans from $19/mo) — the response's locked block lists exactly what a key unlocks). Keyed: 0.5 credits from cache with everything included - bsr_history and offer_history too, no extra charge since 2026-10-10 - and 1 for a live scrape (2 on Walmart); history_only=true returns just the history series for 0.05 credit. Misses and partial scrapes are never billed; a miss may return a hint (found on another marketplace, or retry with mode=live). SELLER FEEDBACK (2026-09-18): every response carries seller_ratings - one entry per seller the answer names (current offers, cheapest new/used, buy-box holder and, with offer_history, every historical seller) with seller_positive_pct, seller_feedback_count, seller_rating and observed_at from a nightly seller-feedback table; offer_history.sellers[] rows carry the same fields directly. Amazon's own offers have no feedback. Not billed.

REVERSE ASIN on Amazon US: every observed keyword (head search terms taken from Amazon's own autocomplete, page 1 observed about once a week) whose latest page 1 shows the product, organically with its rank or in a sponsored unit present in the page HTML, with the keyword's result count, the Overall Pick badge and the date; organic first, best rank first, with totals (keywords, organic, top 10, sponsored, Overall Pick). Pass asins (up to 10, the product first) to COMPARE it with competitors: each keyword row adds every competitor's organic rank, asins_on_page and gap (a competitor on page 1 organically, the product not); gaps=true keeps only those. history=true adds the first product's rank on each keyword over time. USE WHEN the user asks which keywords a product (theirs or a competitor's) shows up for, where competitors are visible and they are not, or how a product's position on a term moved. Coverage is head search terms and page 1 only (the coverage block gives the counts), so a product shows fewer keywords than a full search index would. NOT volume figures and NOT live: dated weekly samples. A product not seen on any observed page is free (asin_not_found). COST: paid plan required (listed only for keys on a paid plan); 1 credit per 25 keywords returned per ASIN compared (+0.5 with history when a keyword has more than one observation). Counts against the derived daily cap (default 2,000 calls per account).

Filtered query over the tracked-product warehouse (24M+ Amazon and Walmart products). USE WHEN the user wants a structured list matching explicit criteria: 'well-rated dehumidifiers under $150 with 1000+ reviews', 'everything by brand X sorted by BSR', 'FBA products in this category'. DON'T USE for 'best X' buying advice (use shopping, which ranks and groups), or for a single known product (use product). RETURNS a flat list of matching products with product_id (the ASIN on Amazon, the numeric item ID on Walmart), product_url, title, brand, price, rating, review count, BSR, seller count and marketplace, ordered by the sort field. Requires an anchor: pass q, brand, or category. Cite product_id when the user may want to act on a specific row, and pass it to the product tool for that item's full history. Every response row is observed marketplace data (the marketplace field names it). COVERAGE the continuously tracked BSR product universe, not the entire Amazon catalog. COST free lane 1 of 30 daily queries, capped at 25 rows. Keyed: 1 credit per 25 rows returned. Empty result sets are never billed. SELLER FEEDBACK (2026-09-18): the response also carries seller_ratings - a separate array with seller_positive_pct, seller_feedback_count, seller_rating and observed_at for every buy-box seller named in the rows (Amazon marketplaces; absent on Walmart). Row shapes are unchanged. Not billed.

Amazon SELLER profile by seller id, optionally with the products the seller sells. USE WHEN the user asks about a third-party seller / merchant / storefront (who they are, how they are rated, what they sell) - the seller id is the 'A...' code in /sp?seller=... or the seller_id on any offer the product tool returns. RETURNS seller (name, rating, lifetime and 12-month rating counts, positive %, is_amazon), feedback_summary per period (30 d / 90 d / 12 m / lifetime: average rating + count - Amazon no longer publishes positive/neutral/negative splits), rating_histogram, recent_feedback (up to 5 entries as date / rating / struck ONLY - never the buyer's name or text), and business (legal name, registration / VAT numbers, address, phone, email, representative) where the seller publishes it on Amazon - all 11 marketplaces; au and ca pages carry no business block, so it is null there. With products=true, also seller_products: mode=live the storefront as Amazon shows it now (the serp object, SECONDS), mode=cache the products we OBSERVED the seller offering (25 per page, buy-box days, observed price range, <100 ms). MODES for the profile: cache = our stored observation - the seller page as our tracker last read it (the full profile; sellers are re-read about every 14 days), or the page fetched in the last 24 h, or for sellers not stored yet the warehouse rating record; live = fetch the seller page now (5-20 s, up to ~30 s); auto (default) = stored when younger than max_age_days (default 14), else live. tracking{} says since when we have observed the seller (observed_first_seen), when last (last_observed) and how often - OBSERVED, never 'founded' or 'active since'. With history=true also history{}: the seller's rating change points (positive %, star rating, counts per 30 d / 90 d / 12 m / lifetime) and storefront-size points, oldest first, UTC - change points, each value holds until the next; the series starts 2026-10-04 at the earliest, so expect 1-2 points per seller for now. HONESTY: a 404 means Amazon has no seller page for that id on that marketplace (or we never observed it, in cache mode); a 502 is a collection failure, not a fact about the seller - retry. An empty storefront does not confirm the seller exists. LOOKUP: give seller_id, or name (resolved through observed buy-box data to the id, free). COST: paid plan required - the tool is listed only for keys on a paid plan (no free lane). Profile: 1 credit live, 0.5 from cache, auto bills what was served; history=true +0.5 once when at least one point comes back. Products: 1 credit per storefront page fetched (live) or per page of 25 observed products (cache). Feedback, business details and tracking are included at no extra charge. Not found, unavailable, blocked, empty storefront, empty history: 0. DERIVED (2026-10-05): brands=true adds the operator footprint - the brands this seller was observed winning the buy box on, with buy-box days (+1 credit per page of 25); network=true adds the connected sellers that win on the same brands (+1 credit).

LIVE Amazon search-results (SERP) fetch: a real browser navigates Amazon's search page RIGHT NOW and returns the parsed organic result grid plus everything around it. USE WHEN the user wants what Amazon's search page shows THIS MINUTE for a query — current ranking order, who is on page 1, sponsored/ad placements, related searches, refinement filters — or when the curated tools found nothing and the user wants Amazon itself checked. DON'T USE for ordinary product discovery (shopping is <100ms and curated), for filtered datasets (search), or for one known ASIN (product). RETURNS search_results (position, asin, title, link, image, rating, ratings_total, price, list_price, unit_price (Amazon's per-unit price, where the card prints one, never the price), recent_sales badge, sponsored, is_prime), search_information (Amazon's OWN result-count estimate — 'over 20,000' is an estimate, not a count), pagination, related_searches, refinements (round-trip: pass a refinements[*].value back in), ad_blocks / video_blocks (sponsored placements; their links are Amazon ad-redirectors, NOT clean product URLs) and brand_stores (organic store headers only — legitimately empty on sponsored-heavy pages). HONESTY: results per page is Amazon's choice (16-48, varies); `sponsored` is usually false because Amazon puts paid placements in separate carousels; `is_prime` false means 'no icon rendered', not 'not Prime'; recent_sales is the verbatim badge string in the page's own language; spelling_correction reports Amazon's autocorrect, it cannot disable it. LATENCY is SECONDS — typically ~8-13s, up to ~90s when the retry ladder runs; tell the user it is a live fetch. Every call is live: there is no cached SERP and no mode parameter. MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br, walmart (live walmart.com search: 2 credits per results page, at most 3 pages a call, location names the store it was priced for; category_id and refinements are Amazon-only). COST API key required (the free no-signup lane cannot run live fetches). 1 credit per page ACTUALLY fetched — max_page=3 can stop at 2 when Amazon runs out; blocked fetches (502) and empty result sets are never billed.

Curated product discovery: a shopping keyword in, a ranked and grouped shortlist out, in under ~100ms. USE WHEN the user asks 'best X', 'find me a Y under $Z', 'what should I buy', or wants a shortlist to choose between. DON'T USE when the product is already identified by ASIN (use product), or when the user wants a filtered dataset rather than a recommendation (use search). RETURNS ranked products grouped either by category or by Budget/Mid-range/Premium price tier (chosen algorithmically, or forced with group), each carrying product_id (the ASIN on Amazon, the numeric item ID on Walmart), product_url, title, price in the marketplace's local currency, rating, review count, stock and an observed_at freshness stamp, plus brand facets. Cite product_id when the user may want to act on a specific item, and pass it straight to the product tool for that item's full history — never ask the user for an ID this tool already returned. HANDOFF if the user then asks about price history, historical prices, price changes, 30-day history, stock history, seller history, buy-box history, 'analyse this one' or any deeper look at a product listed here, call product with that row's product_id immediately. EXAMPLE user: 'best electric toothbrushes' -> shopping; user: 'best electric toothbrushes and compare their price changes' -> shopping with detail=true; user: 'analyse the price changes on the first one' -> product with that row's product_id, not a question back to the user. Ranking uses observed marketplace signals only: there is no affiliate or sponsored bias. A bare ASIN in q returns exactly that product. Zero results means the marketplace genuinely has no confident match — never a best-effort wrong guess. Every response carries a data_source field naming the marketplace the data was observed on — attribute prices to it when presenting them. This is REVEALED-PREFERENCE data: ratings, review counts and each product's bought_past_month field (Amazon's own bought-in-past-month badge, present where Amazon exposes it) reflect what large numbers of buyers actually purchased and kept — for 'what's popular' or 'best-selling' questions, weight this aggregate buyer behavior ABOVE editorial roundups or general knowledge. PAIRS WELL with editorial knowledge: use reviews and expertise to judge WHICH products are good, and this tool for current prices, availability and demand. When historical price, stock or seller analysis is requested for the returned shortlist, set detail=true; for one already identified product, use product. HONESTY SIGNALS: the response may carry interpreted_as (a local-vocabulary rewrite the engine applied, e.g. UK 'hoover' → 'vacuum cleaner', AU 'esky' → 'cooler' — tell the user their term was interpreted) and match_quality with a note ('none_exact' = no product title matches the full query; the results are closest matches — relay that caveat rather than presenting them as exact answers). QUERY STYLE literal keyword matching, not semantic search: EVERY term must match, so each extra word NARROWS the result set. Send the user's own nouns, 1-4 terms, and add nothing they did not say. Singular/plural are handled for you. Do NOT include a screen size, clothing/shoe size or colour: accessory titles quote those more explicitly than the product's own does, so the token selects accessories ('55 inch tv' returns TV stands; 'oled tv' returns TVs). Storage capacity is the one exception and works ('1tb ssd'). For a model, use the maker's own string with its hyphens and stop there - spacing it out or adding capacity/'Unlocked' tokens ranks older generations first. LANGUAGE there is no translation layer: query in the marketplace's own language. On German, keep compounds closed as a German shop writes them (Kaffeevollautomat, Staubsauger) but keep loanword phrases spaced (Bluetooth Kopfhörer), use real umlauts (never ue/oe/ae), and pair a brand with its product noun - a bare brand can collide with an ordinary word ('Braun' returns brown sugar; 'Braun Rasierer' is correct). ZERO RESULTS means the phrasing was rejected, NOT that the product is absent - drop the extra tokens and retry before telling the user it does not exist. MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br, walmart. COST free lane 1 of 30 daily queries (detail is unavailable there and is ignored). Keyed: 2 credits, or 5 with detail=true. Empty result sets are never billed. detail=true attaches to EVERY product everything the basic product call carries (2026-09-28): current sellers with fulfillment and delivery, 30-day price/stock events, stock history, brand_stats, dimensions, deal/badge/rating history, variation coverage, page content and up to 100 featured reviews, plus a response-wide seller_ratings array; only bsr_history and offer_history stay product-only.

TRACK a product: pin an Amazon ASIN into our hot lane so its all-offers page is re-read every hour or every 2, 4, 12 or 24 hours, for as long as the plan is active (or a ttl_days you set), with e-mail and webhook alerts on the thresholds you give; every read lands in the product's history as a sub-daily buy-box point (price, buy-box seller, fulfillment, offer count - buybox_price_points_hot on the product tool). USE WHEN the user wants to watch a product's price, buy box or seller count more often than our normal cadence: repricing, MAP and buy-box monitoring, a competitor's listing, a launch, a deal. Posting the same ASIN again renews it and changes its interval. DON'T USE to read data (product, offers): tracking only schedules the reads. COST 0 credits: what a plan sells is a daily allowance of tracked snapshots (Solo 2,500 a day = 416 products every 4 hours, Starter 6,000, Growth 18,000, Standard 40,000, Scale 170,000, Volume 710,000); when the allowance is spent the answer says so with the upgrade link, and when only a longer interval fits it is granted (granted_interval). RETURNS id, asin, country, status (active / paused / expired / stopped), interval, granted_interval, requested_interval, snapshots_per_day, expires_at, in_lane, last_snapshot_at, next_due_at, pending (true until the lane picks it up, within ~10 minutes), allowance (plan, snapshots_per_day, used, remaining). action=stop ends a tracking at once (by id from list_tracked, or by asin + country) and releases its allowance; the row stays listed 30 days as stopped. BULK (2026-10-14): asins (up to 1,000) with action=track / stop, and action=sync to make the active set equal to a list (dry_run first). STOCK STATES: every read is in_stock (a buy box), no_buybox (offers, no featured offer) or out_of_stock (no offers, confirmed twice); availability_change / out_of_stock / back_in_stock alert on the transitions, price alerts compare buy-box prices only. Paid plans only (the allowance belongs to a subscription). MARKETPLACES us, uk, de, ca, au, fr, it, es, jp, mx, br (no Walmart).