The Currence MCP is live. Ask Claude, ChatGPT, Gemini, or Copilot and get answers backed by Currence data. .

Data Centers & Power

The intelligence layer for the AI-power buildout

Project-by-project intelligence on every data center: power, equipment, participants, milestones, costs, and more.

Learn more

a long exposure of colored lights in the dark

Tracking the AI-power buildout in detail

1,200+

planned US data centers

382GW

announced data center capacity tracked

30,000+

US power projects

project credibility factors scored

Equipment Providers

Find unsupplied projects, size your market, and identify relevant targets with project-level demand signals

Utilities and IPPs

Distinguish credible developers before committing generation, transmission, or PPA capacity

Hyperscalers and Data Center Developers

Benchmark on powering strategy, construction costs, and site selection.

Investors

Underwrite projects with evidence-backed credibility scores, not just announcements

Banks

Assess which projects and opportunities are bankable, with permits, power, equipment, and offtake in hand

See whether the power behind a data center is real

Satellite imagery, moratorium data, and large load tariffs show what stands between an announcement and a project with steel in the ground.

Watch a site move from announced to operational with a dated satellite image timeline tracked on every project with a confirmed location

See where the next project can't go as our moratorium database tracks jurisdiction, restriction type, and coverage area against the projects inside it

Price power before it locks in with a large load tariff tracker following every utility filing through state regulatory review

Deep data on each project

Capacity, participants, and contracts, all attached at the project level, whether you're browsing the platform or pulling it through the API.

See the deal behind the project with supply and offtake contracts attached at the project level: counterparty, contract type, volume, and term

Check the physical build with site and grid detail down to IT capacity, PUE, redundancy tier, rack density, street address, acreage, facility size, and grid operator

Know who's involved with every named participant and their role, from the developer to the power equipment supplier to the anchor tenant

Monarch Compute Campus data page showing details of a data center campus in Point Pleasant, West Virginia, owned by Nscale, announced and last updated 2 September 2026. It lists the sector as Data Centers, estimated costs blurred, capacity blurred, regional average time to power as 60 months, developer tier blurred, power supply on-site, credibility blurred. The summary describes a phased 2,250-acre data center with AI compute capacity and power scalability to over 8 GW, using NVIDIA GPUs and Caterpillar generators, aiming at carbon sequestration and community concerns. Metrics show commercial scale, new build development, 2,250-acre site, and on-site power. Participants include NVIDIA (AI hardware), Microsoft (software), Anthropic (AI research), and Caterpillar (construction equipment), each with company logos and role descriptions.

Data methodology

Every project carries 14 structured fields and an 8-factor credibility score, with status reasoning that captures the why behind every delay or cancellation. The AI data engine continuously pulls updates from news, permits, procurement filings, and regulatory dockets, so the picture stays current as projects move.

Credibility factors

Project status

Development maturity

Project scale

Planned capacity

Developer credibility

Tier 1 to Tier 5, based on track record, financial strength, operational experience, and tenant relationships

Financing signals

Publicly announced funding for project

Power supply arrangements

Executed agreements

Powering plan

Disclosed sourcing strategy

Estimated time to power

Modeled from location, sourcing strategy, interconnection queue timelines, and technology lead times

Commercial offtake

Tenant or hyperscale commitment

Example fields

Location

Address, lat/long

Status

Announced, under construction, delayed, cancelled, operational

Owner

Project owner(s)

Participants

All named project participants

Cost

Announced project cost / investment

Scale

Capacity in MW

Powering model

Grid-connected, off-grid, hybrid

Grid

Utility balancing authority, relevant large load tariff

Powering equipment

Equipment specs

Cooling

Cooling type, suppliers

Supply contracts

Power and equipment agreements

Offtake contracts

Tenant and offtake agreements

Chips

Supplier and quantity

Other metrics

PUE, WUE, Site acreage, Facility square footage

What does Data Centers & Power cover specifically?

The full data center power value chain. We track 1,000+ data center projects and the generation powering them. Coverage includes powering models (grid, behind-the-meter, hybrid), equipment procurement, permit and filing status, and evidence-backed project credibility assessments.

What makes your data center coverage different from what's already out there?

Most data center intelligence products focus on the real estate side: location, capacity, vacancy, leasing. We focus on what's inside the data center and the power infrastructure around it. Nobody else connects the data center buildout to the power market at the project level.

How does the credibility score work?

The credibility score is an evidence-backed assessment of whether a project is likely to actually get built. It considers permit status, equipment procurement progress, developer track record, offtake commitments, and financing signals. The goal is to help you separate real projects from press release pipelines. Credibility scores are live today for all relevant projects (we don't score operational projects) and available via data export.

Who is DC&P built for?

Four primary audiences: utilities evaluating which developers to sign PPAs with, OEMs and equipment providers sizing demand and identifying sales targets, investors underwriting project credibility and allocating capital, and developers benchmarking their power strategy against the competition.

What's included?

Today, you can access project-level data on location, powering models, key participants, and status. As well as:

- Detailed power source and generation technology tracking
- Document feeds with underlying filings and permits
- Detailed project timelines with milestone tracking
- Equipment procurement data (vendor, model, specs, contract status)

Supplier and equipment benchmarks and economics models for project-level cost analysis are coming in Q3 2026.

What geographies do you cover?

We track global data center projects, but are focused on the US market first for extended power data, since this represents the largest and most active data center buildout. We're continuously expanding our international coverage.

Where does the data come from?

We pull from a range of public and proprietary sources, including:

- Regulatory filings (building permits, air quality permits, PUC dockets, FERC filings)
- Interconnection queue data from ISOs and RTOs
- Developer websites and corporate announcements
- Press releases and trade press (Data Center Dynamics, Data Center Knowledge, Utility Dive, etc.)
- Direct validation from market participants (developers, utilities, OEMs, investors)

We ingest this data, enrich it with our AI engine, validate it with our team of expert analysts, and then layer in proprietary intelligence gathered directly from market participants. The enrichment and validation process is what turns raw public filings into structured, actionable project intelligence. The credibility scoring, powering model classification, and equipment tracking are all proprietary to Currence.

Thousands of energy leaders use Currence to finance, build, and operate the energy system.

Get the data