One web data engine. Four ways to use it.

Import.io's four core products are built on shared web-access and extraction infrastructure, while matching, QA, monitoring and delivery capabilities vary by product. Build extractors yourself, use a pricing intelligence application, give AI agents controlled web tools, or have our team run a managed data program. AI datasets such as training corpora, retrieval feeds and evaluation sets are also available as a specialist service.

Shared foundation · product-specific capabilities
Accessreach the web others can't
Renderrun live browser workflows
Extractpages into clean data
Matchavailable where required
OperateQA and delivery by product
JavaScript renderinganti-bot and CAPTCHAdatacenter · regional · residential egressAI schema detectionproduct matching 0.99daily QAprovenance on every record· 500B+ data points a month · since 2012
500B+data points processed per month
2.6Munique products collected per day
5.5B+web pages extracted
2012in production, without interruption, since

Trusted by retailers, brands, marketplaces and the data companies that serve themThree leading commerce-intelligence platforms run their capture on Import.io.

The four surfaces.

Each has its own page with pricing, capabilities and a live demo. This is the map.

Explore integration guides for your existing tools and destinations.

01 · self-service platform

Web data extraction

Paste a URL or describe the data you want. Import.io renders the page, detects the records, proposes a typed schema, finds pagination and publishes an extractor with an API endpoint. Schedule it, deliver it and monitor it. Core extractor workflows are available through the REST API and SDKs as well as the no-code interface.

  • forAnalysts, data teams and developers who want structured web data on a schedule without running a scraping stack
  • you getNo-code and AI extractors · JavaScript rendering · anti-bot and proxy tiers · pagination and chaining · scheduling · change detection · screenshots and HTML · REST API, SDKs, webhooks
  • deliveryCSV, JSON, Parquet to S3, SFTP, webhook; pull from the API
  • pricingFree 30-day trial, 5,000 queries, no card · Standard $199/mo · Professional $399 · Advanced $699, billed annually · full pricing
02 · application

Aperture

Pricing intelligence for retail and brand teams. Competitor prices, promotions, availability and assortment matched SKU by SKU with AI product matching; MAP violations with dated evidence; alerts; an AI market assistant that answers questions about your market with the rows behind the answer.

  • forPricing, category, eCommerce, brand, finance and legal teams at retailers, brands and FMCG companies
  • you getPrice index and competitor matrix · AI product matching with confidence and evidence · MAP case workflow · unauthorized seller detection · alerts · market assistant · exports and API
  • runs onImport.io's hosted web-access and extraction infrastructure; can connect to applicable Import.io extractors and datasets
  • pricingLicensed on tracked SKUs, retailers and cadence, billed annually; matching and onboarding included · talk to sales
03 · for AI agents

Web Scraper MCP

Give Claude, Cursor or any MCP-compatible agent controlled web tools to fetch and render live pages, interact with browser workflows, extract structured data to a schema, and capture HTML or screenshots. Rendering, anti-bot handling and egress run server-side. Setup is usage-based, with a spend cap you control.

  • forDevelopers and AI teams building agents, RAG systems and automations that need the live web
  • you getHosted MCP server at mcp.import.io · rendered page access · structured extraction · rendered HTML inspection · screenshots and browser actions · country targeting · proxy routing and CAPTCHA support
  • works withClaude, Cursor and any MCP client; also available as a REST API
  • pricing10,000 free successful calls, then $0.0002 per successful call; blocked and failed calls never billed · MCP pricing
04 · managed

Managed Services

You describe the data. A named team scopes it, confirms feasibility per source, builds, QA's with AI against the contracted schema, takes your sign-off, then runs it: daily monitoring, alerts in your channel, weekly summary, monthly report, a dashboard you can open yourself, and an SLA with fees at risk.

  • forEnterprises above 500,000 queries a month, heavily protected or authenticated sources, and data companies that want the collection risk off their books
  • you getFeasibility per source in days · contracted schema and acceptance tests · internal AI QA and your UAT gate · scheduled delivery with manifest · monitoring dashboard · named CSM · SLAs on coverage, freshness, correctness and response
  • track record450M queries a year on a single feed · 72 sources under one SLA · three commerce-intelligence platforms buy their capture here
  • pricingPer source maintained, or per query at high volume; fixed onboarding; annual, billed monthly · commercial terms
also · web data for AI

Training corpora, retrieval feeds, evaluation sets

For AI labs and applied teams: domain corpora built to specification with per-document provenance and a quality report, continuous collection for retrieval freshness, point-in-time evaluation sets, and the benchmark that measures all of it.

  • forFoundation model, retrieval and applied AI teams; research groups
  • you getCorpora to spec in Parquet or JSONL with manifests · retrieval change events · frozen evaluation snapshots · sample datasets to inspect · live agent access through MCP
  • pricingPer corpus, per source set, per evaluation set; academic licences available · discuss requirements

Which product do I need?

Start from what you want to end up with. Most customers use two.

you wantWeb data extractionApertureWeb Scraper MCPManaged Services
Structured data from websites, on a schedule, that my team controlsyesnoon demandif you'd rather not run it
Competitor prices, MAP compliance, shelf analytics as answersraw datayesnoas a feed
My AI agents to fetch and extract web data themselvesvia APInoyesno
Sources behind logins or heavy anti-botAdvanced tierincludedpartialyes
More than 500,000 queries a monthnon/aby capyes
A contractual SLA with fees at risknoenterprisenoyes
Someone else to build, QA and maintain itnomatching includednoyes
A training corpus or evaluation set for modelssmall scalenolive accessadditional AI data service
To start today, free, without a callyesdemoyesscoped

The products share web-access and extraction infrastructure, so existing extractors and datasets can often be reused when a workflow moves into Aperture or Managed Services. MCP provides agent-facing access to the hosted rendering and extraction layer. Pricing covers the self-service tiers and MCP; Aperture and Managed are quoted.

What the products build on.

Every surface is built on Import.io's web-access and extraction infrastructure, with additional matching, QA, monitoring and delivery capabilities depending on the product.

  • AccessHosted JavaScript rendering and web access form the core; anti-bot, CAPTCHA, proxy and authenticated-session capabilities vary by product and plan.
  • StructureStructured extraction is shared infrastructure; schema detection, pagination, chaining and document handling are exposed where the product supports them.
  • MatchingIdentifier, attribute and image matching is available in Aperture and applicable managed workflows.
  • QualityValidation, monitoring and acceptance controls are available at different levels depending on the product and service.
  • EvidenceSource HTML and screenshots are available in supported extraction, MCP and managed workflows.
  • GovernanceGovernance controls and service commitments are applied according to the product, plan and managed-data agreement.
companyImport.io, San Francisco · founded 2012 · Connotate acquired 2019
scale500B+ data points/month · 2.6M products/day · 5.5B+ pages extracted
core productsWeb data extraction · Aperture · Web Scraper MCP · Managed Services
additional serviceTraining corpora · retrieval feeds · evaluation sets
accessNo-code · REST API and SDKs · MCP · managed feeds
outputsCSV, JSON, Parquet, HTML, screenshots · API · webhooks · S3, SFTP, warehouses · dashboards
startFree 30-day trial, 5,000 queries, no card · MCP 10,000 free calls · Aperture demo · Managed scoped in days
supportDocs and community · email · chat and training · priority live · named CSM on managed

Product questions.

Short answers. Each product page has the long ones.

What products does Import.io offer?

Four surfaces on one engine: the self-service web data extraction platform (no-code and API), Aperture pricing intelligence, the Web Scraper MCP for AI agents, and Managed Services under SLA. For AI labs, the same infrastructure is offered as training corpora, retrieval feeds and evaluation sets.

Can I move between products without rebuilding?

Often. The products share web-access and extraction infrastructure, so existing extractors and datasets can frequently be reused when moving into Aperture or Managed Services. Product-specific matching, QA, history and delivery features are configured for the chosen workflow.

Which product should I start with?

If you want structured data your team controls, the extraction platform with a free trial. If you want pricing answers rather than rows, Aperture's demo. If you're building agents, the MCP with 10,000 free calls. If the volume is above 500,000 queries a month or the sources are hard, Managed Services.

Can I integrate Import.io data into my existing tools?

Yes. Deliver CSV, JSON or Parquet to S3, SFTP or a webhook, pull from the REST API, or use Aperture's exports into BI tools and pricing engines. Data can be partitioned for Snowflake, BigQuery and Databricks.

Does Import.io support scheduled and on-demand updates?

Yes, depending on the product. The extraction platform and managed services support scheduled workflows, while the API and MCP support on-demand extraction. Aperture refresh cadence is configured for the pricing-intelligence program.

What makes Import.io different from basic scraping tools?

Import.io combines hosted rendering, web access and structured extraction with product-specific options for matching, scheduling, QA, monitoring and delivery. The same foundation supports a pricing application, an MCP server and managed data services.

Pick a surface.
The engine is the same.