Machinepower

Machinepower Index / Dossiers / Brazil

Rank 27 of 50 · Tier 3

Rank 26–31 across the scoring rules tested. How this is tested

Brazil

Diffusion Advantage — For nations that cannot build the frontier but can use it faster than anyone: import the models and win on deployment, across the state and the economy, with public consent.

Score4.75
Potential5.93
Bottleneck cost−1.18
Weakest cell: Hardware & Silicon Access at 2.5 Binding layer: Watts Since the October fix: No change logged Edition: Q4 2026 · updated October 2026

How the Score, Potential and bottleneck cost are worked out: methodology.

The three layers

WattsMaterial power · holds the weakest cell
5.0 / 10

Energy, grids, data centres and silicon — the physical capacity to run models at scale.

Compute tracked by Epoch AI: Brazil controls 1,200 H100-equivalents, wherever the sites are. Hosted compute is shown, not scored.

WeightsIntelligence power
5.2 / 10

Models, alignment capability, standards influence and the talent that produces them.

WillPolitical power
7.6 / 10

The capacity to wire AI into the state and the economy while holding public consent.

Where the score comes from

LayerSub-dimensionBasisScore
WattsPlanning & PermittingAssessed5.6
Firm Power CapacityDataset6.6
Sovereign ComputeDataset5.3
Hardware & Silicon AccessDataset2.5
WeightsFrontier Model OutputAssessed3.4
Alignment CapabilityAssessed4.0
Regulatory StandingDataset7.8
Talent DensityAssessed5.6
WillPolicy CapacityDataset8.8
Public Sector AdoptionDataset9.5
Workforce TransitionDataset5.1
Public TrustDataset6.9

Assessed marks a measurement we judged rather than took from a dataset: 4 of the 12 here.

Sources

Ipsos AI Monitor 2026April 2026
Public Trust in AI
56% say AI products have more benefits than drawbacks

Share of adults agreeing that products and services using AI have more benefits than drawbacks; the Index scores public trust as 1 + share ÷ 10. Survey of 23,532 adults across 32 countries, March to April 2026.

View source →
Ministry of Science, Technology and Innovation (Brazilian AI Plan)2025
Frontier models (assessed)
3.4/10, assessed

The 2024 Brazilian AI Plan funds Portuguese-language models, but domestic model development remains small-scale and well behind the frontier.

View source →
Oxford Insights GAIRI 2025January 2026
Public Sector AI Adoption
PSA 94.78/100 · overall 69.55 (rank 22)

The validated January 2026 report scores Brazil's public-sector adoption at 94.78/100 — among the highest readings anywhere in this Index, and fractionally above the United States' cell as scored (9.5 vs 9.3).

View source →
TOP500.orgJune 2026
High-Performance Compute Capacity
Harpia (Petrobras) — 75.2 PF (#37); ~143 PF national total

Ten Brazilian systems appear on the June 2026 TOP500 with an aggregate Rmax of roughly 143 petaflops, led by Petrobras's Harpia.

View source →
IMF2023 data, published June 2024
AI Preparedness Index: human capital and labour market policies
0.115 of a possible 0.25 · 5.1/10 in the Index

Used for the workforce measurement. The IMF combines schooling, pay, labour mobility and active labour-market policies; the Index rescales it so the weakest of 173 economies scores 0 and the strongest 10.

View source →
Ember2025
Firm generating capacity
161 GW (2025)

Fossil, nuclear, hydro and other dispatchable plant; averaged with generation for firm power.

View source →
GitHub Innovation Graph; Oxford Insights2026
Developer base
8.3 million developers

Second source for talent, averaged with Stanford's count.

View source →
University of Melbourne; KPMGJanuary 2025
Trust in AI at work
60% trust AI at work

Second source for public trust, averaged with Ipsos.

View source →
Oxford InsightsMay 2026
Policy capacity
88/100

Feeds policy capacity.

View source →
Public First, AI Global 20262026
Consent to build data centres
59% want more built, 12% fewer

Half of public trust: support for building more data centres, scored as 1 + share ÷ 10 and averaged with trust in AI.

View source →

Bottlenecks

These hold Brazil below its Potential. Its two lowest scores cost it the most: hardware & silicon access at 2.5 (part of Watts) and frontier model output at 3.4 (part of Weights).

Watts (Material)Hardware & Silicon Access

Assessment

The great deployment state of the Global South: government uptake of AI is the strongest cell in Brazil's profile — scored fractionally above even the United States' — and its hydropower grid is one of the cleanest large systems on Earth. What it lacks is the layer beneath: no leading-edge fabs, modest sovereign compute, and a frontier yet to produce a globally notable model.

The art of the possible

Convert deployment excellence into gravity: use continental-scale procurement and clean hydropower to anchor regional inference capacity, then bargain hosting into technology transfer. Brazil does not need to build the frontier to matter — it needs to become the place the frontier runs for half a billion people.

Strategic play — Diffusion Advantage

Turn deployment into gravity: use continental-scale procurement and clean hydropower to anchor inference for half a billion people, and bargain hosting into technology transfer. Brazil does not need the frontier to matter; it needs the frontier to run on Brazilian power.

Leverage — the nearest position. No chokepoint in the chain. Its leverage is the grid — clean firm power at a scale most of the chain's holders cannot match — and the proof, useful to every middle power, that a state can deploy AI it does not own.

Open Brazil in the live Index → — every source behind every number, and the comparison tool.