Yes. The Free tier is 600 requests a month at 100 requests a minute, across every endpoint in the catalog, with no credit card. Nothing about it expires.

What happens when I hit the quota?

Requests return 429 with the limit, what you have used, and which window you tripped — nothing is queued, and nothing is charged. The monthly quota resets on the 1st (UTC). There is no rollover: unused requests do not carry into the next month.

Do failed requests use up my quota?

A 5xx does not. If a source breaks on our side, the request is excluded from your monthly count. A 4xx does count — a wrong parameter is still a request we answered. Per-minute throttling counts everything, so a retry loop is still throttled.

Which tier do I need?

Start on Free. Every response carries the remaining quota in its headers, and the dashboard reads the same number enforcement uses, so you will see the ceiling coming rather than discovering it.

How is Zapi different from RapidAPI or Apify?

Zapi focuses on a consistent REST contract — one API key, one response format, one error pattern for every source. No need to learn per-provider quirks like on traditional marketplaces, and no need to set up scraping workflows like on Apify.

What about the legality of accessing public data?

Zapi only accesses data that is publicly available without a login. Responsibility for how it is ultimately used stays with the developer — make sure your use case complies with the original source's ToS and the regulations in your operating region.

Is there an official SDK?

One: zpi-sdk, for TypeScript and JavaScript. It carries a typed error hierarchy, retries only what is safe to retry, submits bulk jobs, and verifies webhook signatures. Everything else is plain REST — cURL or any HTTP client works.

Can I request a source that isn't in the catalog yet?

Yes, through the community Discord, on any tier.

Why are Pro and Ultra marked Soon?

Checkout is not open yet. The limits and features listed on those tiers are already built and enforced — what is pending is billing, not the product.

arXiv

Searches an open-access repository of scientific preprints and returns paper metadata.

Endpoints: 1
Requests: 10
Updated: 3 months ago
For LLMs: llms.txt

Endpoint health17 hours ago

searchpassing, 1.1s

1/1 passingdaily 01:00 WIB

Playground

1 endpoint

GET/v1/search-tools:arxiv/search5m cache3 paramschecking key…

Parameters

Response

EXAMPLE

{
  "page": 1,
  "count": 5,
  "items": [
    {
      "url": "http://arxiv.org/abs/2201.00978v1",
      "title": "PyramidTNT: Improved Transformer-in-Transformer Baselines with Pyramid Architecture",
      "pdfUrl": "https://arxiv.org/pdf/2201.00978v1",
      "source": "arxiv.org",
      "authors": [
        "Kai Han",
        "Jianyuan Guo",
        "Yehui Tang"
      ],
      "summary": "Transformer networks have achieved great progress for computer vision tasks. Transformer-in-Transformer (TNT) architecture utilizes inner transformer and outer transformer to extra…",
      "publishedAt": "2022-01-04T04:56:57Z"
    },
    {
      "url": "http://arxiv.org/abs/2104.11502v1",
      "title": "Learning to Cluster Faces via Transformer",
      "pdfUrl": "https://arxiv.org/pdf/2104.11502v1",
      "source": "arxiv.org",
      "authors": [
        "Jinxing Ye",
        "Xioajiang Peng",
        "Baigui Sun"
      ],
      "summary": "Face clustering is a useful tool for applications like automatic face annotation and retrieval. The main challenge is that it is difficult to cluster images from the same identity …",
      "publishedAt": "2021-04-23T09:43:36Z"
    },
    {
      "url": "http://arxiv.org/abs/2404.05657v1",
      "title": "MLP Can Be A Good Transformer Learner",
      "pdfUrl": "https://arxiv.org/pdf/2404.05657v1",
      "source": "arxiv.org",
      "authors": [
        "Sihao Lin",
        "Pumeng Lyu",
        "Dongrui Liu"
      ],
      "summary": "Self-attention mechanism is the key of the Transformer but often criticized for its computation demands. Previous token pruning works motivate their methods from the view of comput…",
      "publishedAt": "2024-04-08T16:40:15Z"
    }
  ],
  "total": 174551,
  "hasMore": true,
  "nextPage": 2
}

Code examples

zpi-sdkrecommended

// npm i zpi-sdk
import { ZpiClient } from "zpi-sdk";

const client = new ZpiClient({ apiKey: "zpi_xxxxxxxxxxxxxxxxxxxxxxxx" });

const data = await client.run("search-tools:arxiv", "search", {
  "query": "transformer",
  "count": 10,
  "page": 1
});
console.log(data);

Full SDK reference →

cURL

request.sh

curl -X GET "https://api.zpi.web.id/v1/search-tools:arxiv/search?query=transformer&count=10&page=1" \
  -H "x-api-key: zpi_xxxxxxxxxxxxxxxxxxxxxxxx"

Full reference

1 endpoint · plain text

GETSearch/v1/search-tools:arxiv/searchReturns matching preprints with titles, authors, abstracts, and links for a query.

Parameters

querystringrequired: Search keyword
countnumberoptional: Results per page 1-30 (default 10)
pagenumberoptional: Result page (default 1, max 50)

Request

curl -X GET "https://api.zpi.web.id/v1/search-tools:arxiv/search?query=transformer&count=10&page=1" \
  -H "x-api-key: zpi_xxxxxxxxxxxxxxxxxxxxxxxx"

Example response

{
  "page": 1,
  "count": 5,
  "items": [
    {
      "url": "http://arxiv.org/abs/2201.00978v1",
      "title": "PyramidTNT: Improved Transformer-in-Transformer Baselines with Pyramid Architecture",
      "pdfUrl": "https://arxiv.org/pdf/2201.00978v1",
      "source": "arxiv.org",
      "authors": [
        "Kai Han",
        "Jianyuan Guo",
        "Yehui Tang"
      ],
      "summary": "Transformer networks have achieved great progress for computer vision tasks. Transformer-in-Transformer (TNT) architecture utilizes inner transformer and outer transformer to extra…",
      "publishedAt": "2022-01-04T04:56:57Z"
    },
    {
      "url": "http://arxiv.org/abs/2104.11502v1",
      "title": "Learning to Cluster Faces via Transformer",
      "pdfUrl": "https://arxiv.org/pdf/2104.11502v1",
      "source": "arxiv.org",
      "authors": [
        "Jinxing Ye",
        "Xioajiang Peng",
        "Baigui Sun"
      ],
      "summary": "Face clustering is a useful tool for applications like automatic face annotation and retrieval. The main challenge is that it is difficult to cluster images from the same identity …",
      "publishedAt": "2021-04-23T09:43:36Z"
    },
    {
      "url": "http://arxiv.org/abs/2404.05657v1",
      "title": "MLP Can Be A Good Transformer Learner",
      "pdfUrl": "https://arxiv.org/pdf/2404.05657v1",
      "source": "arxiv.org",
      "authors": [
        "Sihao Lin",
        "Pumeng Lyu",
        "Dongrui Liu"
      ],
      "summary": "Self-attention mechanism is the key of the Transformer but often criticized for its computation demands. Previous token pruning works motivate their methods from the view of comput…",
      "publishedAt": "2024-04-08T16:40:15Z"
    }
  ],
  "total": 174551,
  "hasMore": true,
  "nextPage": 2
}

{ "page": 1, "count": 5, "items": [ { "url": "http://arxiv.org/abs/2201.00978v1", "title": "PyramidTNT: Improved Transformer-in-Transformer Baselines with Pyramid Architecture", "pdfUrl": "https://arxiv.org/pdf/2201.00978v1", "source": "arxiv.org", "authors": [ "Kai Han", "Jianyuan Guo", "Yehui Tang" ], "summary": "Transformer networks have achieved great progress for computer vision tasks. Transformer-in-Transformer (TNT) architecture utilizes inner transformer and outer transformer to extra…", "publishedAt": "2022-01-04T04:56:57Z" }, { "url": "http://arxiv.org/abs/2104.11502v1", "title": "Learning to Cluster Faces via Transformer", "pdfUrl": "https://arxiv.org/pdf/2104.11502v1", "source": "arxiv.org", "authors": [ "Jinxing Ye", "Xioajiang Peng", "Baigui Sun" ], "summary": "Face clustering is a useful tool for applications like automatic face annotation and retrieval. The main challenge is that it is difficult to cluster images from the same identity …", "publishedAt": "2021-04-23T09:43:36Z" }, { "url": "http://arxiv.org/abs/2404.05657v1", "title": "MLP Can Be A Good Transformer Learner", "pdfUrl": "https://arxiv.org/pdf/2404.05657v1", "source": "arxiv.org", "authors": [ "Sihao Lin", "Pumeng Lyu", "Dongrui Liu" ], "summary": "Self-attention mechanism is the key of the Transformer but often criticized for its computation demands. Previous token pruning works motivate their methods from the view of comput…", "publishedAt": "2024-04-08T16:40:15Z" } ], "total": 174551, "hasMore": true, "nextPage": 2 }

// npm i zpi-sdk import { ZpiClient } from "zpi-sdk"; const client = new ZpiClient({ apiKey: "zpi_xxxxxxxxxxxxxxxxxxxxxxxx" }); const data = await client.run("search-tools:arxiv", "search", { "query": "transformer", "count": 10, "page": 1 }); console.log(data);

Playground

Code examples

Full reference

Parameters

Request

Example response

Related APIs

Brave Search

Bing Search

DuckDuckGo

GitHub Search

Wikipedia

Hacker News

Playground

Code examples

Full reference

Parameters

Request

Example response

Related APIs

Brave Search

Bing Search

DuckDuckGo

GitHub Search

Wikipedia

Hacker News