An X-Prize for AI in Africa

Africa AI
Prize

AI inference is the new electricity. But most of the world can’t afford to switch it on.

A $10M prize to transform Africa with highly capable, open & affordable AI inference.

Scroll to explore

Make AI work at
African economics.

Frontier intelligence is already here, in the air, everywhere, like current with nowhere to plug in. The barrier is not capability or geography. It is cost. AI retails at Silicon Valley rates in economies where that locks out 95% of potential users. If intelligence were cheap, it would already be everywhere on this continent, the way mobile money is.

Left unsolved, the cost wall compounds. AI adoption is a relative velocity race: every year the price stays up, infrastructure deficits multiply, talent migrates, capital flows elsewhere. What is a 20-year gap today becomes a 50-year gap by 2035. The continent that does not build its own AI capacity will have AI built for it, on terms it did not negotiate.

The stakes reach well beyond this continent. More and more of human progress and value creation now routes through AI, in every economy on Earth, and any society that does not get competent with it falls behind in productivity and grows relatively poorer. Cheap inference is no magic bullet. A thousand other things still have to go right. But get access to intelligence wrong, and in the competition of nations the rest may increasingly not matter.

"If you do not control the intelligence layer, you do not control the economy it reshapes."

20yr

Current AI gap

50yr

Projected by 2035

1.4B

People at stake

54

African nations benefit

$10M in prizes.
Open-source.
Built for Africa.

The Enzi Prize targets a 50-100x reduction in the cost of AI inference: from roughly ₦550 per million tokens today to ₦5.5-11. Two tracks: Local AI (models running on the hardware people actually own, phone first) and API-Served (radically cheaper serving on reference hardware). Solutions must be open source, reproducible, and royalty-free for African use. Healthcare, agriculture, education, and governance are where the impact lands, not categories of entry.

This is an investment in sovereign capability. A $14M initiative designed to catalyze an ecosystem, and extricate from AI dependency and rent.

"We are not asking for permission to participate in the AI revolution. We are building the infrastructure to lead it."

$10M

Total prize pool for teams building sovereign AI solutions for the continent

50-100x

Target cost reduction in AI deployment across African infrastructure

$2M+

Research scholarships and fellowships to keep leaderboard teams in the fight

Open Source

All solutions must be open source, reproducible, and royalty-free for African/Global South use

Track 1 · Local AI

Models running on the hardware people actually own, phone first · $300k

Track 2 · API-Served

Radically cheaper serving on reference hardware · up to $5M Platinum

Why a prize works

A well-designed prize mobilizes many times its purse in private R&D: the $10M Ansari X-Prize drew over $100M in team spending and seeded the commercial spaceflight industry. We define the verified output and pay only on results; teams and their backers supply the rest.

Registered teams get access to a subsidized shared GPU pool and a continuous public benchmark leaderboard.

New to prize competitions? Read how a prize competition works.

The Fellowship Fund

Prizes have a known failure mode. A team gets on the leaderboard, proves the approach works, then runs out of money eighteen months before the finish line. The result is a worse outcome and smaller competitive field, and the teams still standing are the ones who may already be funded.

At least $2M is set aside to stop that. Any team at threshold on the public leaderboard can draw on the Fellowship Fund: research scholarships and fellowships that cover the people, not the prototype. Compute will come from a shared GPU pool. This covers the salaries and stipends that keep researchers on the problem.

The fund is aimed squarely at diaspora researchers and African academics. There are thousands of them in labs and universities outside the continent working on adjacent problems, and the reason they are not working on this one is rarely interest. It is that they cannot afford to. This pays for it.

Eligibility follows the leaderboard. Application details will be published with the challenge briefs.

"The prize picks the winner. The fellowship decides how many teams can contribute to the outcome."
"This may not be won by billions of dollars and marquee labs. It may be won by hundreds, then thousands, of people attacking the problem from every angle. The prize exists to point both well-funded labs and independent tinkerers (and every research lab in between) at one target."

Q2 2026

Official announcement and challenge categories revealed

Q4 2026

Official launch at Moonshot, Lagos · October 28 · teams begin building

Lagos.
October 28, 2026.

Not just another panel discussion. A working conference for practitioners: engineers, researchers, founders, and operators who are building AI for and from Africa. Presented in partnership with Tech Cabal as part of the Moonshot Conference series.

6 stage sessions. An invitation-only roundtable. A speakers and government mixer the night before.

See the full programme →

Location National Theatre, Lagos
Date October 28, 2026
Format Tech Cabal Moonshot Conference
Audience Practitioners only. Builders, not observers.

A coalition for
sovereign AI.

Africa Deep Tech Foundation

Grassroots Arm

Tech Cabal

Media & Conference Partner

Future Africa

Ecosystem Partner

Talent Nation

Talent Partner

Build with us.

The Enzi Prize is assembling teams, sponsors, and participants who believe Africa's AI future must be built by Africans. Register your interest below.

Partners and teams can also email us directly at hello@enziprize.org