The Model Index
World/Africa
AI access, cost and use · Southern Africa · Africa · data read 1 Oct 2026

Namibia

In Namibia, a $20-a-month AI plan costs 22% of the median living standard, 27 times the share it takes in the United States; for 76% of people it is over a tenth of what they live on. Claude, OpenAI's API and Gemini are all officially offered. 65% of people use the internet, and Claude use per working-age person is 0.44 times the world average.

22%of the median living standard for a $20 planAfrica median 38% · United States 0.79%
76%of people for whom the plan is over a tenth of their living standardabout 2.3 m people
3of 3leading services officially offeredClaude · OpenAI · Gemini
65%of people use the internetworld 74% · 3.1 m people

The figuresNamibia

People3.1 m
Median living standard$1,111a year per person · consumption
A $20 plan takes22%of that · 27× the US share
For whom it is over a tenth76%of people
Survey2015revalued at 2025 prices
GNI per person$4,340plan is 5.5% of it
Claude use per person0.44×world average = 1
Notable AI models built here0

Officially offered here3 of 3

Run it yourself →
Claude · claude.ai appYes
OpenAI · APIYes
Gemini · web appYes

From each lab's own published list: Anthropic's Claude.ai countries, OpenAI's API countries and Google's Gemini web app countries. Not listed means not officially served; open-weight models can be run anywhere.

Against Africa and the worldregional figures are medians or population-weighted

MeasureNamibiaAfricaWorld
A $20 plan ÷ median living standard22%38%9.9%
People for whom it is over a tenth76%96%63%
Plan price ÷ GNI per person5.5%14%2.9%
People using the internet65%40%74%
Claude use per working-age person0.44×0.24×1.00×
Notable AI models built001,256

What it is used forNamibia against the world

Report ↗
Namibia
Work 36%
Personal 28%
Coursework 36%
The world
Work 42%
Personal 41%
Coursework 17%
Share of the world's Claude use0.02%0.04% of the world's people
Work36%world 42%
Personal28%world 41%
Coursework36%world 17%
Handed over to the model (automation)53%world 48%

Anthropic Economic Index, May 2026: Claude.ai conversations by country. One product from one lab, so read it as a signal of who is using frontier models, not a census.

What people make with itshare of conversations

Document or report20%
Explanation or answer14%
Advice or recommendation6%
Math or calculation6%

The most common things people in Namibia ask Claude to produce.

NeighboursSouthern Africa · lightest burden first

CountryPeopleOnline$20 plan ÷ median living standardOver a tenthClaude use per personOfficial lists
South Africa65 m78%19%73%0.46×3 of 3
Namibia3.1 m65%22%76%0.44×3 of 3
Botswana2.6 m57%24%79%0.52×3 of 3
Lesotho2.4 m52%44%96%not reported3 of 3
Eswatini1.3 m63%no estimate–not reported3 of 3
All countries →

The foundationsNamibia against the world

World Bank ↗
People using the internet
65%world 74%
People with electricity
59%world 92%
Mobile subscriptions per 100 people
85world 112
Fixed broadband lines per 100 people
2.8world 20.1
Enrolled in university-level education
26%world 44%
Research spending, share of GDP
0.65%world 2.6%
Researchers per million people
225world 1,652
Secure internet servers per million people
427world 16,381

The latest World Bank figure for each measure. The tick marks the world figure.

A $20-a-month plan ($240 a year, US list price) against the median household living standard per person in each country. The median is from the latest national household survey (World Bank Poverty and Inequality Platform), revalued from 2021 purchasing-power dollars into US dollars at recent prices and official exchange rates, assuming real living standards have not changed since the survey. Some surveys measure income and others consumption, which includes food people grow and housing they own, so this is a living standard, not cash to spare. Countries with no survey since 2015 or no defensible exchange rate are left blank.

World Bank household surveys ↗Anthropic countries ↗OpenAI countries ↗Gemini countries ↗Anthropic Economic Index ↗Epoch AI notable models ↗