Diffbot is a set of APIs that reads the public web and returns structured data instead of raw HTML or markdown. Its Extract API classifies a single URL and returns structured JSON. Its Crawl product walks an entire site from one seed URL and collects pages into a single dataset. Natural Language resolves entities, relationships, facts and sentiment from plain text. The Knowledge Graph API runs structured queries against a graph of companies, people and news from the public web. Web Search is a low-latency search API that the page says is free of ads and telemetry, and it can be self-hosted.
The site is aimed at businesses and developers who need web data in a form their software can use directly. The pages show sales briefs, financial due diligence and network mapping as example workflows, and a LeadGraph product is listed in the navigation.
The pricing page lists a Free plan with 10,000 credits a month and no credit card required. Paid plans are a Startup plan at $299 a month with 250,000 credits, a Plus plan at $899 a month with 1,000,000 credits, and a custom Enterprise plan. Each action uses a set number of credits. Extracting a page costs 1 credit, and exporting a Knowledge Graph entity costs 25. Usage above a plan's monthly credits is billed per credit. Plans are monthly and can be cancelled at any time. Students and academic researchers can get Startup-tier access free through Diffbot for Students.
The documentation is available on the site, and sales questions can be sent to sales@diffbot.com.
