scrapethissite.com with an AI agent
The short answer: scrapethissite.com is a sandbox site for practising web scraping, with tables, forms and paging. We gave Claude Code 2 real tasks there, each in a brand-new session. With Playwright MCP a task used 210K tokens on average. With Dejo it used 188K the first time and 80K on every run after, 62% fewer, because Dejo replays a learned task without the model reading the pages again.
The tasks
- “find the hockey team stats for the Mighty Ducks of Anaheim and tell me how many goals against they had in the 1997 season”
- “go to page 5 of the hockey team table and tell me the name of the first team listed on that page”
What each tool used
| task | Dejo, first run | Dejo, every run after | Playwright MCP | Playwright CLI |
|---|---|---|---|---|
| 1 | 209K tokens, 21 s | 79K tokens, 1 call, 10 s, right | 250K tokens, 4 calls, 24 s, wrong | 445K tokens, 7 calls, wrong |
| 2 | 168K tokens, 16 s | 82K tokens, 1 call, 11 s, right | 169K tokens, 2 calls, 16 s, right | 496K tokens, 8 calls, right |
Calls are browser tool calls the agent made. Right or wrong is checked against the answer read off the live site before any tool ran.
Learned and replayed with Dejo on 13 November 2026
We ran the task again with Dejo on 13 November 2026, so you can see what it saves today. These are the steps Dejo kept, exactly as it reported them. Each one carries a check that it landed, and a replay walks them in plain code.
1. go https://www.scrapethissite.com/pages/
2. click link "Hockey Teams: Forms, Searching and Pagination" (by name)
3. fill label "Search for Teams" (by name)
4. press Enter
5. answer (by position)Replay: 5 steps in 3.4 s, with no page reading. Answer: “261” (right).
A new task, once Dejo knows the site
In our site-memory test, the task “go to page 5 of the hockey team table and tell me the name of the first team listed on that page” ran twice in fresh sessions: once with an empty memory, and once after Dejo had learned a different task on scrapethissite.com. It used 174K tokens the first way (4 turns) and 175K the second (4 turns), right both times.
How we measured
Claude Sonnet 5 at medium effort, a brand-new Claude Code session per task and tool, @playwright/mcp 0.0.80, October 2026. The full method, and the sites that blocked every tool, are on the benchmark page. Raw data: benchmark-real-2026-10-21.json, benchmark-real-2026-10-23.json, benchmark-real-2026-10-23-fixes.json, benchmark-site-memory-2026-11-10.json.
Other websites we measured
- packages.debian.org: Debian's package directory, searched by name and release
- rfc-editor.org: the official home of the RFC series of internet standards
- catalog.ccv.edu: the course catalog of the Community College of Vermont
- ctan.org: the Comprehensive TeX Archive Network, where LaTeX packages live
- practicesoftwaretesting.com: a demo tool shop built for practising test automation
- demowebshop.tricentis.com: Tricentis's demo web shop for test automation
- collection.internationalfolkart.org: the online collection of the Museum of International Folk Art in Santa Fe
- gutenberg.org: Project Gutenberg, a library of free ebooks
- pub.dev: the package site for Dart and Flutter
- hackage.haskell.org: the Haskell community's package archive
FAQ
How many tokens does an AI agent use on scrapethissite.com?
In our test, Claude Code with Playwright MCP used 210K tokens per task on scrapethissite.com on average. Dejo used 188K the first time and 80K on every run after, 62% fewer than Playwright MCP once the task was learned.
Can Claude Code automate scrapethissite.com?
Yes. In fresh Claude Code sessions on 2 tasks there, Dejo answered 2 of 2 right, Playwright MCP 1 and Playwright CLI 1. With Dejo, a task done once is replayed later without the model reading the pages again.
Does a new task on scrapethissite.com get cheaper once Dejo knows the site?
In our site-memory test, a new task there used 174K tokens with an empty memory and 175K after Dejo had learned a different task on the same site (4 turns against 4).