“Free” doesn’t mean the same thing twice in the scraper API market. Some vendors grant credits per month, some hand you a one-time pile that never refills, some count every request while others count credits per page — and some ask for your credit card before you’ve seen a single byte of data. If you’re evaluating a free Amazon scraper API, the number that matters isn’t the headline figure. It’s three things: does the grant recur, does it require a card, and is the data Amazon-specific or generic HTML extraction.
Most “best free API” roundups never touch the APIs they rank. We did. On October 16, 2026, we registered a fresh account on every platform on this list and ran the same live test against each one: the same product (ASIN B0FQFB8FMG, Apple AirPods Pro 3) on the same marketplace (amazon.com). For each platform we recorded the signup process, the actual free-tier terms we received, the API response time, the response format, the field count, and whether the data was correct. Two of the five failed outright. The numbers below are measured, not copied from pricing pages.
The comparison (tested 2026-10-16)
| Provider | Free tier (as tested) | Card required? | Live test result | Response | Best for |
|---|---|---|---|---|---|
| Pangolinfo (our product) | 60 credits/month, resets monthly | No | ✅ Product: 51 fields in 8.5s; Reviews: 4,318 total, newest 2 days old | Structured JSON | Deepest Amazon data on a free tier |
| Chocodata | 1,000 requests, one-time | No | ✅ 60 fields in 3.3s, price/rating/review count all correct | Structured JSON | Fastest structured Amazon product data |
| ScraperAPI | 5,000 credits, 7-day trial | No | ✅ 1.8 MB raw HTML in 4.2s — no CAPTCHA, but no parsing either | Raw HTML | Raw page fetches at volume |
| Bright Data | Advertised 5,000/month | No | ❌ New account showed $0 balance and “suspended”; API timed out at 120s | — | Not testable on the free tier as advertised |
| APISerpent | Free calls on signup | No | ❌ HTTP 502 on both attempts (9.7s / 16.8s) | — | Could not be tested — endpoint down |
1. Pangolinfo — the deepest Amazon data behind a free tier (our product)
Disclosure first: this is our product, and it’s listed first for that reason. The free tier is 60 credits every month, reset monthly, no credit card required. Product pages cost 1 credit per page; review pages cost 5 credits per page — so 60 credits is 60 product pages, or 12 review pages, or a mix every month, on an ongoing basis.
Tested: 51 product fields, 4,318 reviews, 6 credits total
Our test burn: one product call (1 credit, 8.5s) returned 51 structured fields — title, price ($199.00), 16,512 ratings, 7 images, variants, seller info, Best Sellers Rank, and more. One review call (5 credits, 11.5s) returned page 1 of 4,318 total reviews, with the newest review dated two days before the test — genuinely fresh data, not a cached snapshot. Total cost of this entire evaluation: 6 credits out of the 60 you get every month.
Amazon focus: objects, not HTML
Generic scraper APIs return whatever HTML they fetched and leave parsing to you. Pangolinfo’s parsers are built around Amazon objects: product detail pages, keyword search results (with paid and organic positions separated), category product mix and density, seller storefronts and assortments, Best Sellers rankings, New Releases, delivery-time estimates, and full review pages via the dedicated Amazon Review API — 20+ Amazon-specific endpoints, not one generic fetcher pointed at amazon.com. Coverage spans 20+ marketplaces. The difference is between “we fetched a page” and “we understand what an Amazon page contains.”
Freshness and success rate
Amazon data goes stale fast — prices move, reviews land daily. The review test above returned content two days old, pulled live. The published success rate is 99%, and anti-bot handling (CAPTCHAs, IP rotation, fingerprinting) is server-side: your code never sees it. When a free tier is your evaluation environment, what you’re really evaluating is whether the data is real and current. Ours was.
Three ways in: REST, MCP, Skill
The same free key unlocks three integration surfaces. The REST API for any language or stack. The Amazon Data MCP server — 20 tools that let Claude, Cursor, and other AI agents call live Amazon data mid-conversation, no glue code. And the Scraper Skill for agent IDEs like OpenClaw. If your end goal is feeding Amazon data to an AI agent rather than downloading CSVs, the free tier is the on-ramp to all three — a breadth none of the generic APIs on this list match.
2. Chocodata — the strongest competitor we tested
Chocodata earned second place honestly: it worked, and it worked well. Signup was email + password with a 6-digit email code, no card. The free tier is a one-time grant of 1,000 requests (5 credits per request, billed only on success). Our test hit their Amazon product endpoint and got HTTP 200 in 3.3 seconds — 7.3 KB of structured JSON with 60 fields: title, price ($199, correct), rating 4.4, 16,512 reviews, buybox info, variations, sales rank, rating distribution, images, bullet points. The catch is the word “one-time”: 1,000 requests that never refill. Excellent for a serious evaluation; not a tier you can run on.
3. ScraperAPI — biggest free grant, but you parse the HTML yourself
ScraperAPI’s trial is genuinely generous: 5,000 credits for 7 days, no card (we registered and verified this). Our test fetched the same Amazon product page in 4.2 seconds — and returned 1.8 MB of raw HTML, HTTP 200, no CAPTCHA. The title and price were in there. But so was everything else, unparsed: no JSON, no fields, no structure. If you need raw page fetches at volume and own a parser, this is the largest free grant on the list. If you need Amazon data ready to use, budget the parsing work — that’s the real cost of “free” here.
4. Bright Data — advertised free tier did not materialize
This one surprised us. Bright Data advertises 5,000 free credits per month with no credit card required, and signup indeed asked for no card — email, password, and a 6-digit email code. But the new account showed Balance $0, plan “Pay as you go,” and an account status of “suspended due to reached negative balance.” The advertised free credits never arrived. Our API test against their Web Unlocker endpoint timed out at 120 seconds with no response. We’re reporting exactly what happened: as of October 2026, a fresh signup could not use the service at all without adding funds. Check their current terms — but verify the credits actually land before you plan around them.
5. APISerpent — endpoint down during testing
APISerpent’s signup is Google OAuth only (no email/password option), which we completed. Their docs describe a 35-field structured Amazon product endpoint at $0.020 per 1,000 requests. Our test called it twice with the same ASIN: HTTP 502 after 16.8s, then HTTP 502 after 9.7s. The endpoint was down for the duration of our testing window, so we couldn’t evaluate data quality at all. We’re leaving them on the list because the offering is real — but “tested” has to mean something, and what we measured was an outage.
What the tests actually prove
Three patterns emerged. First, “free” without usable data is the most expensive option — Bright Data’s $0-balance suspension and APISerpent’s 502s cost us more time than any credit limit. Second, raw HTML is not Amazon data — ScraperAPI’s 1.8 MB response is only useful after you build the parser Pangolinfo and Chocodata already run. Third, Amazon depth is the differentiator: of the two APIs that returned
structured data, only Pangolinfo pairs it with a dedicated review API (4,318 paginated reviews), 20+ Amazon-specific endpoints, and AI-agent
surfaces (MCP + Skill). Chocodata is a strong general scraper with good Amazon coverage; Pangolinfo is built for Amazon first.
One more honest note about free tiers. Whatever the size, a free allowance is enough for hobbyists — and it’s enough for a professional team to run a real evaluation. But for production workloads, no free tier on this list is going to carry you. That’s not a flaw; it’s simply not what free tiers are for. The real value of a usable free tier is that it drops the evaluation barrier to zero: your data team can test against real ASINs, real marketplaces, and real freshness requirements before anyone has to sign anything.
How to start free with Pangolinfo
Three steps, no credit card:
- Sign up. Open pangolinfo.com and click “Start Free Trial” — this takes you to tool.pangolinfo.com to create your account. The 60-credit monthly free tier is active immediately.
- Copy your API key. Sign in to the tool.pangolinfo.com dashboard and copy your API key from the API key page. Use a long-lived key and never paste it into shared chat text.
- Make your first call. The fastest path is the API playground in your dashboard, or a single REST request following the quickstart in the docs. A product-page lookup costs 1 credit — you can pull 60 pages before the month is out.
Prefer to start from code? The companion GitHub repo has runnable Python and Node.js examples that hit the free tier: Pangolin-spg/amazon-scraper.
Enterprise user? Contact our support team for personalized testing and a tailored demo of the Amazon Scraper API.
FAQ
Is the free tier really free?
Yes. No credit card is required, and the 60 credits reset every month. When you run out mid-month, calls pause until the reset — or until you upgrade.
What can 60 credits actually do?
More than it sounds for evaluation: 60 product detail pages, or 12 full review pages, or a realistic mix (e.g., 30 product pages + 6 review pages). Our own test of this article burned 6 credits total. That’s enough to verify data quality, field coverage, and freshness against your real ASINs before spending anything.
When should I upgrade?
When free stops being enough: scheduled daily monitoring, higher page volumes, or team use. The paid tiers start at $19/month (Starter). Until then, the free tier is a complete evaluation environment, not a demo.
Testing methodology: all five platforms were tested on October 16, 2026 with fresh accounts, using ASIN B0FQFB8FMG (Apple AirPods Pro 3) on amazon.com. Response times, field counts, and account states are as measured on that date. Free-tier terms change — verify against each vendor’s current page before deciding.
