UNESCO education statistics in your AI assistant: example questions

MCP server for the UNESCO Institute for Statistics: ~5,000 indicators on education, science, culture and communication, by country, region and year. Connect any MCP client to https://uis.sidneybissoli.com/mcp — nothing to install, no account, no API key — and ask in your own words. The questions below were run against the live server on 2026-10-13; the figures are from UIS data release 20260507-91260335 and a later release can revise them.

Example questions

Enrolment: “How did tertiary enrollment grow in Brazil, Mexico and Chile since 2015?”

Em português: “Como evoluiu a matrícula no ensino superior no Brasil, no México e no Chile desde 2015?”

What the assistant calls:

uis_search_indicators {"query": "gross enrollment ratio tertiary"} — "enrollment" is searched as the UIS's "enrolment", and the response says so
uis_get_data {"indicators": ["GER.5T8"], "geo_units": ["BRA", "MEX", "CHL"], "start_year": 2015}

What came back: Gross enrolment ratio for tertiary education (%): Brazil 49.1 (2015) → 69.7 (2024); Mexico 32.0 → 48.2; Chile 86.4 → 110.2 — every value flagged as a national estimate.

Literacy: “What is the youth literacy rate in India, Nigeria and Pakistan, for girls and boys?”

Em português: “Qual a taxa de alfabetização dos jovens na Índia, na Nigéria e no Paquistão, por sexo?”

What the assistant calls:

uis_search_indicators {"query": "youth literacy rate"}
uis_get_data {"indicators": ["LR.AG15T24.F", "LR.AG15T24.M"], "geo_units": ["IND", "NGA", "PAK"], "start_year": 2015}

What came back: Youth literacy rate, 15–24 years (%), latest year available: India 96.6 female / 98.0 male (2024); Nigeria 81.4 / 81.3 (2024); Pakistan 66.5 / 79.5 (2021). Years differ by country — the answer lists what exists instead of filling gaps.

Education spending: “How much do Brazil, Argentina and South Africa spend on education as a share of GDP?”

Em português: “Quanto Brasil, Argentina e África do Sul gastam com educação, em % do PIB?”

What the assistant calls:

uis_search_indicators {"query": "government education spending GDP"} — "spending" is searched as "expenditure"
uis_get_data {"indicators": ["XGDP.FSGOV"], "geo_units": ["BRA", "ARG", "ZAF"], "start_year": 2015}

What came back: Government expenditure on education as a percentage of GDP (%), latest year: Brazil 5.6 (2022), Argentina 5.0 (2023), South Africa 6.0 (2024).

Every answer also carries a provenance block: the exact UIS Data API URL, the data release, the retrieval time, the licence (CC BY-SA 4.0) and a ready-to-paste citation.

How it compares

Measured on 2026-10-13. There is no dedicated “Apify UNESCO UIS MCP server”: Apify offers a community scraper actor for the UIS, which its general MCP server can expose as a tool.

Tool	What it is	Access and cost	Prefer it when
uis-mcp-server (this server)	Hosted MCP server over the official UIS Data API: catalogue search with everyday-wording translation, data by country/region/year, the UIS data release pinned on every query and a provenance block (source URL, release, retrieval time, licence) on every answer.	Free, no account, no API key — point the client at the URL. Also on npm (stdio).	The question is asked in an assistant and the answer has to be auditable.
UNESCO UIS Statistics Scraper (Apify actor, thirdwatch)	Community actor on Apify that reads the same UIS Data API and returns rows with source URL, qualifier and footnotes; reachable as an MCP tool through Apify's general MCP server (mcp.apify.com). The data release is not stated in the output. Not a dedicated MCP server: one actor, inputs are indicator codes, geo units and years — no catalogue search.	Apify account (OAuth or API token); pay per result — US$ 0.004 on the free plan down to US$ 0.002 on paid plans, plus US$ 0.00005 per run (Apify API, 2026-10-13).	You already run Apify pipelines and want UIS rows in an Apify dataset.
unesco-mcp (Python, PyPI)	Community MCP server over the UIS API, run locally with uvx; indicator catalogue cached in SQLite; more analysis tools (rankings, geography comparisons).	Local install (Python, uvx). Free, no key.	You prefer a local server and ranking/comparison tools built in.
global-education-mcp (Python)	Community MCP server combining the UIS Data API with the OECD Education at a Glance API.	Local install (Python). Free.	You want UIS and OECD education data side by side in one local server.
uisapi (R) / unesco-reader (Python)	Community client libraries of the UIS Data API, for analysis code rather than an assistant.	Install from CRAN / PyPI. Free.	You are writing R or Python and want UIS series in a data frame.

The full comparison, including World Bank clients and the UIS bulk files, is in the README.

Independent project — not affiliated with or endorsed by UNESCO or the UNESCO Institute for Statistics. Contact: sbissoli76@gmail.com