DataForSEO API

DataForSEO is a provider of SEO data delivered through APIs. Instead of exporting reports from a dashboard, you send a request (for example, “the Google results for this keyword in the UK”) and get structured JSON back, which you can analyse in Python, Google Sheets or a database. MLforSEO uses it throughout its courses and guides as the main source for SERP and keyword data.

How it works

The API is split into groups, each with many endpoints:

• SERP API: live or queued results pages from Google and other engines, with every element typed (organic results, People Also Ask, video, AI Overviews and more).
• Keywords Data API: search volume, CPC, competition and trends from sources such as Google Ads and Google Trends.
• DataForSEO Labs API: its own keyword and domain databases, used for keyword ideas, keyword difficulty, search intent, ranked keywords and competitor discovery.
• Backlinks, OnPage, Content Analysis, Merchant, App Data and Business Data APIs.
• AI Optimization API: data on brand mentions and keyword demand in LLM-based assistants.

Pricing is per request on a prepaid balance, so cost scales with how many keywords, locations and features you pull. There is also a no-code option for simple SERP pulls.

How it’s used in SEO and marketing

• Collecting SERPs on a schedule to build your own rank and SERP-feature tracker.
• Pulling the ranking pages for a keyword set and analysing their entities, n-grams and formats.
• Classifying intent from the SERP itself (which features and page types appear).
• Building keyword universes with volume, difficulty and intent in one dataset.
• Tracking AI Overview and AI Mode citations alongside organic rankings.

Example

In MLforSEO’s guide to automating SERP analysis, DataForSEO supplies three endpoints: Organic SERP (advanced), AI Mode and Historical SERPs. The guide notes that you need to set load_async_ai_overview to true to capture AI Overviews that are not cached, which DataForSEO charges an extra $0.002 for and refunds when no AI Overview appears, and that historical snapshots cost $0.00012 each (prices as of October 2026). The sample dataset cost a few cents of credit, and the code wraps requests in a spend cap so a loop cannot drain the balance.

Limitations and things to watch

• Raw responses are deeply nested; plan the parsing into tidy tables before scaling up.
• Costs are small per call but add up across locations, devices and daily schedules. Check the pricing page before big runs.
• Search volumes and difficulty scores are estimates and will differ from other tools.
• SERPs are personalised and volatile, so treat a single pull as a snapshot and track changes over time.

Learn it on MLforSEO

• 10 Ways to Automate SERP Analysis with AI
• What Is SERP Feature Analysis?
• Identifying Desired Content Formats & Platforms from SERP Data (Lab Notebook)
• Semantic AI-powered SEO Keyword Research course, SERP collection lab