Overview

Nimble is best known as web search and retrieval for AI agents. The same infrastructure that lets agents search and browse the live web also powers a complete web scraping toolkit. Teams use it to pull structured data from specific sites, run recurring collection at scale, or hand the entire pipeline to Nimble and receive data in their warehouse.

There are three ways to scrape with Nimble:

Option	Best for	You get
Extract API (plus Crawl and Map)	Any URL, with full control over rendering, actions, and parsing	Markdown, HTML, screenshots, or typed fields from a parsing schema
Extract Template API	Popular sites where you want normalized data without writing selectors	Production-ready JSON, maintained and auto-healed by Nimble
Managed data services	Teams that want the whole web data pipeline run for them	Clean, scheduled data streamed into Snowflake, Databricks, S3, and more

Extract API: any URL, full control

Extract turns any URL into clean data. Use it when you need precise control over how a page is retrieved and parsed.

• Output formats: markdown, raw HTML, full-page screenshots, HTTP headers, and structured fields from a parsing schema (CSS selectors or schema-based parsing)

• JavaScript rendering for single-page apps and client-loaded content (React, Vue, Angular)

• Browser actions such as click, scroll, fill, and wait, to reach content behind "load more" buttons, filters, tabs, and infinite scroll

• Network capture to return the data a page loads from its internal APIs

• Localization by country, state, or city for local prices, inventory, and results

• Reliable access to protected and JS-heavy sites, handled automatically with stealth browsing and a residential network, with no proxy setup on your side

• Batch and async requests, with webhook and cloud storage delivery

Crawl retrieves whole sections of a site in one request, and Map discovers every URL on a domain so you can plan collection.

Price: $1 per 1,000 requests for Extract, Crawl, and Map. Extract docs

Extract Template API: maintained scrapers for popular sites

Extraction Templates are pre-built, production-ready extractors. Pass a template name and parameters, such as a product ID, search term, or URL, and get back normalized JSON with consistent fields. There are no CSS selectors to write and no maintenance: Nimble maintains each template around the clock and auto-heals it when the target site changes.

Coverage includes:

• E-commerce: Amazon product pages (including regional Amazon sites), Walmart, Target search, and more

• Search engines: Google Search results pages, including organic results, ads, and local packs

• AI search and answer engines: Google AI Overviews, Google AI Mode, ChatGPT, and Gemini, including the sources each answer cites

• Local and maps: Google Maps search and Google Maps reviews

• Social and other sources: TikTok and a growing catalog of community templates across news, patents, travel, and jobs

Capabilities:

• Single requests or batches of up to 1,000 items

• On-demand or scheduled runs

• Pagination for search results and listings

• Localization for location-specific pricing and availability

• Delivery by polling, webhooks, or directly to S3 or GCS

Price: $3 per 1,000 requests. A Managed Service by Nimble option, where Nimble's experts configure and manage template-based delivery for you, is available for a 10% fee. Extraction Templates docs

Example: e-commerce data

Source	Primary identifier	Typical fields
Amazon	ASIN	title, price, list price, currency, availability, rating, review count, brand, category, images
Walmart	SKU / UPC	price, was price, availability, pickup and delivery, badges, rating
Target	TCIN	product name, price, brand, images, category, availability

Common uses: competitive and dynamic pricing, MAP monitoring, digital shelf analytics, and assortment tracking.

Example: search and AI visibility

Run the Google Search, AI Overviews, AI Mode, ChatGPT, and Gemini templates on a schedule across a keyword or prompt set, by location and device. Track organic and paid positions, local pack presence, whether an AI answer appears, and which domains it cites. Load the results into Power BI, Looker Studio, or your warehouse to report on visibility across classic and AI search.

Managed data services: Nimble runs the entire pipeline

For teams that want data, not infrastructure, Nimble's experts configure and operate the whole web data pipeline: source selection, collection, parsing, normalization, quality checks, scheduling, and delivery. Data streams continuously into the customer's stack.

What Nimble handles:

• Configuring collection for your sources, products, keywords, and locations

• Custom data normalization and enrichment

• Product matching across different sources

• Schema enforcement, deduplication, anomaly detection, and PII masking

• Scheduling, from real-time to hourly, daily, or custom cadences

• Ongoing maintenance as sites change

Delivery: streamed to Snowflake, Databricks, BigQuery, S3/GCS, Azure, or via API and webhooks, in JSON, CSV, or Parquet.

Plans (billed annually):

Plan	Price	Includes
Startup	$2,500/month	5 concurrent agents, 350K monthly web page credits, 7-day data storage, custom agent ETL, MCP integration
Scale	$7,000/month	10 concurrent agents, 1.2M monthly web page credits, 30-day data storage, custom agent ETL
Professional	$15,000/month	Higher concurrency and credits, longer storage, custom workflows
Enterprise	Custom	Unlimited agents, custom storage, advanced security and SLAs

Talk to sales to scope a managed pipeline.

When to use search or Web Search Agents instead

Scraping works best when you know which sites and pages you need. When you don't:

• Use Search to find the right pages first, then Extract or a Template to retrieve them.

• Use Web Search Agents when the job spans many unknown sources, such as building a dataset of every retailer selling a product, enriching a list of companies, or producing a cited research brief. The agent plans, searches, browses, extracts, and returns schema-matched output with citations.

Proof points

• Extraction reliability: 90% extraction completeness and 99.1% fetch success across 5,856 URLs and 2,107 domains, from an open, reproducible benchmark. extract-benchmarks

• Grips Intelligence: improved accuracy and stability across 45,000 sites and 65,000 brands. Case study

• Alta: >99% job success at millions-per-day scale. Case study

Compliance

Nimble collects public web data with compliance by design: SOC 2 Type II, GDPR and CCPA alignment, DPA availability, ethical IP sourcing with external legal audits, and a headers mechanism for website operators. Zero Data Retention is available on enterprise plans. Trust Center

Pricing summary

• Free tier: 5,000 requests per month, access to all APIs, no credit card required

• Extract, Crawl, and Map: $1 per 1,000 requests

• Extract Template API: $3 per 1,000 requests (+10% for Managed Service by Nimble)

• Media API: $2 per GB for images, video, audio, and documents

• Managed data services: from $2,500/month

• Volume discounts: available through sales

See nimbleway.com/pricing.