Tooling benchmarks

A compact readout of OSINT-oriented tasks. Browser scores require completing a search workflow and returning target evidence; output volume is not treated as quality.

3Categories

11Cases

50Evidence probes

Category readout

Coverage by task category.

Category	Cases	Coverage	Evidence
Browser Automation	4	50%	partial
PDF Extraction / OCR	3	98%	complete
Scraping	4	53%	partial

Browser Automation

Open the task brief to inspect sources, search flows, and expected evidence.

Task brief4 investigative cases

Companies House registry search to company filing history

Load the UK Companies House register, type OPENAI into the search form, open OPENAI UK LTD, and navigate to its filing history so a reporter can inspect downloadable filings.

URL: https://find-and-update.company-information.service.gov.uk/
Search: OPENAI

must contain: OPENAI UK LTDmust contain: 14367667must contain: Filing historymust contain: View PDFmust contain: Confirmation statement

OpenSanctions name screening search

Load OpenSanctions, search for a named person, and return the sanctions/PEP evidence for the matching entity.

URL: https://www.opensanctions.org/search/
Search: Roman Abramovich

must contain: Roman Abramovichmust contain: Personmust contain: PEPmust contain: Sanctionedmust contain: Russia

OpenStreetMap place search for a municipality

Load OpenStreetMap, search for Zug Switzerland, and return the Nominatim result context and relation links so a reporter can verify the place and map target.

URL: https://www.openstreetmap.org/
Search: Zug Switzerland

must contain: Search Resultsmust contain: Nominatimmust contain: Zug, Schweiz/Suisse/Svizzera/Svizramust contain: relation/1686447must contain: relation/1683151

Wikidata entity search for structured company identity

Search Wikidata for OpenAI, open the matching entity record, and return structured identity evidence useful for company/person disambiguation.

URL: https://www.wikidata.org/wiki/Special:Search
Search: OpenAI

must contain: OpenAImust contain: Q21708200must contain: American artificial intelligence research organizationmust contain: openai.commust contain: instance of

dev-browser

100%

Playwright script

100%

browser-use terminal

browser-harness

Tool	Evidence	Coverage	Speed	Spend	Note
dev-browser	complete	100%	9,578 ms	n/a	executed
Playwright script	complete	100%	6,591 ms	n/a	executed
browser-use terminal	missed	0%	4,042 ms	n/a	no target evidence returned
browser-harness	missed	0%	30 ms	n/a	CDP keepalive ping timeout

PDF Extraction / OCR

Open the task brief to inspect sources, search flows, and expected evidence.

Task brief3 investigative cases

Follow the Money investigative manual with reordered cover/front matter

Parse a known investigative PDF from local cache, with a source URL available for remote parsers.

URL: https://cps.ceu.edu/sites/cps.ceu.edu/files/cps-joint-publication-follow-the-money-2004.pdf

must contain: Follow the Moneymust contain: International Budget Projectmust contain: resource cursemust contain: Open Society Institutemin chars: 20000

GIJN Citizen Investigations guide with OSINT topic coverage

Parse a known investigative PDF from local cache, with a source URL available for remote parsers.

URL: https://gijn.org/wp-content/uploads/2019/09/Citizens_Investigation_Guide.pdf
Page range: 0-5

must contain: Citizen Investigationsmust contain: Bellingcatmust contain: Finding out who owns corporationsmust contain: Digging up property recordsmin chars: 50000

UNESCO Story-Based Inquiry manual

Parse a known investigative PDF from local cache, with a source URL available for remote parsers.

URL: https://www.media-diversity.org/additional-files/documents/A%20Guides/A%20manual%20for%20investigative%20journalists%20[EN]%20[FR]/A%20manual%20for%20investigative%20journalists%20[EN].pdf

must contain: Story-Based Inquirymust contain: investigative journalismmust contain: hypothesismin chars: 40000

LlamaParse

100%

Poppler pdftotext baseline

100%

Docling

100%

Fireparse / Firecrawl document parse

100%

Extend Parse 2.0

100%

Surya OCR

83%

LangExtract over extracted text

83%

Tool	Evidence	Coverage	Speed	Spend	Note
LlamaParse	complete	100%	130,751 ms	n/a	executed
Poppler pdftotext baseline	complete	100%	125 ms	n/a	executed
Docling	complete	100%	27,450 ms	n/a	executed
Fireparse / Firecrawl document parse	complete	100%	18,407 ms	n/a	executed
Extend Parse 2.0	complete	100%	31,995 ms	426 credits	executed
Surya OCR	complete	83%	194,577 ms	n/a	executed
LangExtract over extracted text	complete	83%	68,114 ms	n/a	executed

Scraping

Open the task brief to inspect sources, search flows, and expected evidence.

Task brief4 investigative cases

Basel-Stadt council protocol listing

Retrieve a civic-protocol listing and preserve links or labels that would let a reporter reach meeting records.

URL: https://grosserrat.bs.ch/ratsbetrieb/ratsprotokolle?all=1

must contain: Grosser Ratmust contain: Ratsprotokollemust contain: PDFmin chars: 1000

Companies House direct registry-result scrape

Retrieve a public registry result page and preserve the target company name, number, status/date, and address context.

URL: https://find-and-update.company-information.service.gov.uk/search/companies?q=OPENAI

must contain: OPENAI UK LTDmust contain: 14367667must contain: Incorporated on 21 September 2022must contain: 50 Broadway

Lausanne Conseil communal minutes/PV page

Retrieve a French-language council minutes page and preserve the minutes/PV source trail.

URL: https://www.lausanne.ch/officiel/autorites/conseil-communal/seances-et-pv.html

must contain: Conseil communalmust contain: séancesmust contain: PVmin chars: 1000

Zurich Gemeinderat protocol listing

Retrieve a multilingual civic-protocol page and preserve meeting/protocol evidence.

URL: https://www.gemeinderat-zuerich.ch/protokolle

must contain: Gemeinderatmust contain: Protokollmust contain: Sitzungmin chars: 1000

Firecrawl scrape

75%

Obscura headless browser

46%

Exa contents API

38%

Tool	Evidence	Coverage	Speed	Spend	Note
Firecrawl scrape	complete	75%	1,242 ms	n/a	executed
Obscura headless browser	partial	46%	1,032 ms	n/a	no target evidence returned
Exa contents API	partial	38%	4,022 ms	$0.0020	no Exa contents results returned