Search Data Hub content with a phrase and filters, then return matching page records with source metadata and content text.
Web Search Rerank
Search the web from a natural-language request and return relevance-ranked webpage results with content and source metadata
Aperçu
Web research often starts with a question rather than a known page. This Data App turns a natural-language request into a relevance-ranked set of webpages, preserving the page text, concise summaries, source identity, and publication context needed for analysis.
It is useful for quickly building a research set around a company, product, topic, or event. Freshness and domain filters help focus the search, while account, organization, and search-role filters can narrow the results when those source attributes are available.
Data notes
Each row represents one webpage returned for the submitted search question. Page URL is used to identify a result and remove duplicates. The result set covers webpage records only; a separate image collection returned by the source is not included here.
Page content may be empty even when a title, snippet, or summary is available. Account name, organization name, site icon, cached URL, language, and family-friendly or navigational flags may also be unavailable for individual pages. Publication and crawl timestamps are preserved as source text without inferring a timezone.
What the results look like
Each row is one relevance-ranked webpage result.
| Page URL | Page title | Site name | Published at | Snippet |
|---|---|---|---|---|
| https://www.example.com/article/123 | Xiaomi car delivery momentum continues to rise | Example Site | 2026-08-05 14:23:10 | Xiaomi car deliveries have continued to grow. |
Additional values include the full page content, summary, display and cached URLs, account and organization names, language, source identifiers, and crawl time.
Use cases
- Build a research set for a company, product, market, or event by combining Page title, Page content, Snippet, and Summary.
- Monitor recent coverage by pairing a Freshness filter with Published at and Last crawled at.
- Compare source visibility by grouping results by Site name, Account name, or Organization name.
- Preserve traceable source links by using Page URL and Display URL when reviewing or exporting results.
Périmètre et limites
Each run accepts one natural-language search request and returns up to 200 relevance-ranked webpage records in one response. Separate image results are not included.
- Adapté pour
- When you need relevance-ranked webpage results for a natural-language research question
- When you need webpage titles, summaries, content, source metadata, and publication timestamps in a structured result set
- When you want to narrow web search by freshness, domains, account, organization, or search role
- À ne pas utiliser pour
- When you need image-only results or a dedicated image search dataset
- When you need private, login-only, or non-public webpage content
Gestion des échecs
Comportement en cas d'échec et de nouvelle tentative déclaré par le développeur. Nous vous recommandons de l'inclure dans votre prompt système lors de l'intégration.
- 1If no useful webpages are returned, broaden the query or remove restrictive domain and freshness filters
- 2Retry later after rate limiting, temporary service unavailability, or a timeout
- 3If authentication fails, verify the API Key configured for the execution environment
Entrée
Paramètres requis pour appeler cette App, générés à partir de l'input.schema du manifest.json.
- Search question
queryRequisstringThe natural-language topic or question to search. Provide at least one character; clearer wording produces more focused results.ExempleXiaomi car delivery and sales performance - Result count
countFacultatifintegerThe number of webpage results to request. The default is 10 and the allowed range is 1 to 200; fewer results may be available for a narrow query.Plage1–200Défaut10 - Freshness filter
freshnessFacultatifstringLimits results by recency. Use noLimit, oneDay, oneWeek, oneMonth, oneYear, a date such as 2026-08-05, or a date range such as 2026-08-01..2026-08-31. The default is noLimit.DéfautnoLimit - Included domains
includeFacultatifstringSearch only the listed domains. Separate up to 100 domain values with | or commas, using domain values supplied by the Data Hub domain list. - Excluded domains
excludeFacultatifstringExclude the listed domains from the search. Separate up to 100 domain values with | or commas, using domain values supplied by the Data Hub domain list. - Account name
account_nameFacultatifstringFilters results to an exact account name when the source identifies one. Leave empty to search without an account filter. - Organization name
office_nameFacultatifstringFilters results to an exact organization name when the source identifies one. Leave empty to search without an organization filter. - Search role
role_codeFacultatifstringUses a configured Data Hub search role, such as WEB_SEARCH. Leave empty to use the system default role.
Sortie
Structure des champs d'un enregistrement, générée à partir de l'output.schema du manifest.json.
- Account name
account_namestringThe account name associated with the webpage when available.ExempleChina News Network - Cached page URL
cached_page_urlstringThe cached webpage URL when available. - Page content
contentstringThe webpage body content when available. It is independent from the snippet and summary and may be empty.ExempleXiaomi car deliveries have continued to grow, with stronger showroom traffic and steady delivery scheduling. - Last crawled at
date_last_crawledstringThe source timestamp of the latest crawl, preserved as returned without timezone conversion.Exemple2026-08-05 15:00:00
Exemple de données
Généré à partir d'output.schema. Les champs sans valeur d'exemple utilisent des espaces réservés structurels.
{
"account_name": "China News Network",
"cached_page_url": "<cached_page_url>",
"content": "Xiaomi car deliveries have continued to grow, with stronger showroom traffic and steady delivery scheduling.",
"date_last_crawled": "2026-08-05 15:00:00",
"date_published": "2026-08-05 14:23:10",
"display_url": "https://www.example.com/article/123",
"id": "0",
"is_family_friendly": false,
"is_navigational": false,
"language": "en",
"name": "Xiaomi car delivery momentum continues to rise",
"office_name": "<office_name>",
"site_icon": "<site_icon>",
"site_name": "Example Site",
"snippet": "Xiaomi car deliveries have continued to grow.",
"summary": "Xiaomi car deliveries have continued to grow, with stronger showroom traffic and steady delivery scheduling.",
"url": "https://www.example.com/article/123"
}Intégration
Cette App peut être intégrée via MCP, API, SDK ou export de fichiers — tous les canaux partagent les mêmes capacités et la même tarification. Chaque requête s'authentifie avec l'en-tête Authorization: Bearer à l'aide d'une API Key (longue durée, créée dans la console Data Hub) ; les clients MCP peuvent aussi se connecter via OAuth, sans clé. D'autres options comme la CLI et Skill arrivent bientôt.
Avec MCP (Model Context Protocol), vous pouvez appeler cette App directement depuis des clients IA comme Claude et Cursor. Choisissez votre client et votre mode d'authentification, puis copiez la config ci-dessous.
Config du client
Remplacez la valeur après Bearer par votre API Key longue durée. Fonctionne dans n'importe quel client, en CI ou en environnement sans navigateur.
{
"mcpServers": {
"octoparse_datahub": {
"type": "http",
"url": "https://mcp-v2.octoparse.com?pin=meme_today/data-hub-web-search-rerank",
"headers": { "Authorization": "Bearer <YOUR_API_KEY>" }
}
}
}Laissez l'IA tout configurer pour vous
Vous ne voulez pas modifier les configs à la main ? Copiez le prompt d'installation et collez-le dans n'importe quel client IA — il effectuera la configuration à sa manière. (Par défaut, il vous guide pour vous connecter et ne vous demande jamais d'envoyer votre API Key dans la conversation.)
detail.access.mcp.composeHint
Tarifs
Facturé selon le nombre d'enregistrements renvoyés avec succès. Les exécutions échouées ne sont pas facturées. Facturé par unités de 50 enregistrements ; toute unité incomplète est arrondie à 50 enregistrements.
Plusieurs événements de facturation se cumulent indépendamment. Voir chaque élément pour les détails. Les exécutions échouées ne sont pas facturées.
Voir les créditsEssayer maintenant
Renseignez les paramètres et lancez l'exécution — les résultats proviennent d'un appel réel.