The one-sentence answer
If "more compute" means announced, OpenAI wins by a mile (9+ GW). If it means online and serving customers today, Google DeepMind and Anthropic lead, because almost nobody's headline gigawatts are actually plugged in yet.
The gap between those two sentences is the whole story.
The honest reframe
You asked who has more compute, scoped to three things that matter for delivering consistent service: data-center power (GW), installed accelerator fleet, and peak inference-serving capacity. Here is what the research kept running into:
Nearly every headline number is contracted, not operational. OpenAI's 9 GW, Anthropic's 5 GW AWS envelope, the 3.5 GW Broadcom deal: all signed, almost none online. The only firmly operational figures the verification gate would accept were OpenAI's ~0.3 GW at Abilene (independently confirmed by satellite imagery) and xAI's ~300 MW Colossus 1 as of mid-2025. Everything else is a press release with a future date attached.
So the ranking that matters for your question (consistent top-level service) is not the announced-capacity leaderboard. It is the much shorter list of who has mature compute online, serving real traffic, with enough platform diversity to absorb a demand spike without melting down.
03 · The rankingReliability-weighted ranking (for consistent service delivery)
| Rank | Lab | Why | Confidence |
|---|---|---|---|
| 1 (tie) | Google DeepMind | Most vertically integrated: owns the silicon (TPU), the data center, and the network. TPUv7 Ironwood is at near-parity with Nvidia's flagship. Only player proven at coherent multi-datacenter training. Most genuinely-online mature capacity. | High |
| 1 (tie) | Anthropic | Most platform-diversified live fleet (1M+ Trainium2 in use now, 1 GW+ of Google TPU landing in 2026, plus Nvidia GPUs). Diversity is a reliability hedge no single-platform rival has. Caveat below. | High |
| 3 | OpenAI | Largest future ceiling (9+ GW, 2M+ chips) but the most constrained present: only ~0.3 GW online against the largest user base in the industry. Structurally the most exposed to the compute crunch today. | High |
| 4 | xAI | Held the largest single fully-operational coherent cluster (Colossus 1, mid-2025). But its 2026 expansion claims failed verification, and it is the least diversified on serving. | Medium |
| Unranked | Meta, DeepSeek, Mistral | Their compute claims either failed adversarial verification (Meta's Prometheus/Hyperion, DeepSeek's fleet) or never surfaced (Mistral). This is an evidence gap, not a confirmed absence of compute. | Low |
The reliability caveat that cuts against the obvious read: on June 2, 2026, six days before this report, Claude suffered a major global outage that Anthropic attributed to "unexpected capacity constraints." Anthropic has also run documented usage limits on Claude through 2026. Diversification helps, but demand is outrunning supply even for the best-hedged lab. No lab in this set is immune to capacity-driven throttling right now.
04 · Axis 1: PowerAxis 1: Data-center power (GW)
The bottleneck of the era is not chips, it is power. Here is online-versus-announced, the only distinction that matters.
| Lab | Backer / source | Online now (mid-2026) | Contracted / announced | Reality check |
|---|---|---|---|---|
| OpenAI | Microsoft + Oracle + Stargate/SoftBank | ~0.3 GW (Abilene only; 6 other sites at 0 GW) | >9 GW by 2029 (≈20M H100-equiv of compute) | ~97% of the headline is not built. Abilene itself capped ~1.2 GW per March 2026 reporting. |
| Anthropic | AWS (primary) + Google Cloud TPU | >1 GW TPU landing in 2026, plus ~1 GW of the AWS envelope by end-2026 | 5 GW AWS + up to 1M TPU + 3.5 GW Broadcom (2027+) | Best near-term online ratio of the group. ~80% of AWS capacity is still forward. |
| Google DeepMind | Google Cloud TPU v7 | Large mature TPU fleet (no single GW figure disclosed; serves Google-scale traffic today) | 43-superpod, ~400K-chip fabric | Least announced-hype, most actually running. |
| xAI | Self-built Colossus (Memphis) | ~0.3 GW (Colossus 1, confirmed mid-2025) | Colossus 2 / ~2 GW / 555K GPUs | Forward 2026 claims failed verification. Confirmed figure is a 2025 snapshot. |
| Meta | Self-built | Unverified | Prometheus ~1 GW (2026), Hyperion up to 5 GW | Both cluster claims failed verification. |
Reading the gigawatts (sizing-data-centers lens): the "9 GW ≈ 20 million H100s" line is a compute-equivalent measure, not a physical chip count. Do the arithmetic and 9 GW across 20M H100s implies ~450W per chip, well below an H100's real ~1kW all-in draw. The reconciliation: Blackwell and Rubin parts deliver far more FLOPs per watt, so "20M H100-equivalents" of compute fits into far fewer, far more powerful physical chips. Translate GW to fleet at a realistic ~1.4 kW/accelerator (chip plus PUE overhead of ~1.25-1.35 for these liquid-cooled builds) and 1 GW of IT load is roughly 700K-1M modern accelerators. That is the unit to keep straight when a lab quotes you a gigawatt number.
05 · Axis 2: FleetAxis 2: Installed fleet and largest coherent cluster
A single coherent cluster (one fabric, low-latency, trains one model) is worth more than the same chip count scattered across regions. Here is what survived verification.
| Lab | Largest coherent cluster (verified) | Fleet notes |
|---|---|---|
| xAI | Colossus 1: ~200K H100/H200 + ~30K GB200, ~300 MW | Largest fully operational, single-coherent cluster at the mid-2025 snapshot. |
| Anthropic | Project Rainier: ~500K Trainium2 across multiple US data centers | 1M+ Trainium2 chips in use now to train and serve Claude; plus up to 1M TPUv7 Ironwood incoming. |
| Google DeepMind | 9,216-TPU coherent superpod (1.77 PB shared HBM, 9.6 Tb/s interconnect); scales to ~400K chips across 43 superpods via Jupiter networking | The only player demonstrated at coherent multi-datacenter training. |
| OpenAI | Abilene GB200 buildout (capped ~1.2 GW) | 2M+ chips under development across the Stargate program. |
| Meta | Unverified | Cluster claims failed verification. |
On raw per-chip maturity, Google's TPUv7 Ironwood lands within ~10% of Nvidia's GB200/Blackwell on peak FLOPs, memory, and bandwidth (4.6 PFLOPS dense FP8, 192 GB HBM3e, 7.4 TB/s). That matters because Google's TPU fleet underpins two labs' serving at once: its own (DeepMind) and Anthropic's.
06 · Axis 3: ServingAxis 3: Peak inference-serving capacity (the axis that maps to your purpose)
This is the soft axis, and the research is honest about why: no measured serving metric survived verification. Not tokens/sec, not concurrent-user ceilings, not p99 latency, not a documented throttling incident, for any lab. The serving-reliability comparison therefore rests on three proxies:
- Capacity online in 2026 (not announced for 2029). Favors Google DeepMind and Anthropic.
- Platform diversification (can you reroute around a failed platform?). Strongly favors Anthropic (Trainium + TPU + Nvidia). Google is single-platform but owns that platform end to end.
- Known incidents. The confirmed signal here is the June 2, 2026 Claude outage and Anthropic's 2026 usage limits, both pointing to demand outrunning supply industry-wide. Reporting also flags a broader 2026 compute crunch with rationing and rising GPU prices across the sector.
Net read for consistent service: Google DeepMind's vertical integration (silicon to serving) and Anthropic's platform diversity are the two structurally strongest positions. OpenAI carries the most serving risk today because it pairs the largest demand with the smallest online footprint. None of them is currently serving without strain.
07 · AttributionPer-lab attribution (who is actually behind each "lab")
Since these are pure labs, the compute is really their backers' compute:
- OpenAI → Microsoft + Oracle + Stargate/SoftBank. Headline scale, thin operational base. The 4.5 GW Oracle deal (reported >$300B over 5 years, "reported, not fully confirmed") brings the program past 5 GW under development.
- Anthropic → AWS as primary training partner (Project Rainier, 1M+ Trainium2, up to 5 GW) plus Google Cloud TPU (1 GW+ online in 2026, up to 1M chips) plus a 3.5 GW Broadcom TPU deal from 2027. The only genuinely multi-platform lab.
- Google DeepMind → Google Cloud TPU v7, fully owned and integrated. No external backer to coordinate with.
- xAI → self-built Colossus. Confirmed operational lead in mid-2025, least-verified forward trajectory.
- Meta → self-built Prometheus/Hyperion. Claims unverified.
What did not survive verification (and why that matters)
Eight claims were killed in 3-vote adversarial verification. Treat these as unconfirmed, not disproven, but do not build a ranking on them:
- Meta Prometheus as a 1 GW cluster online in 2026 (0-0).
- Meta Hyperion scaling to 5 GW (1-0).
- xAI Colossus 2 sized for ~110K GB200 NVL72 / gas-turbine power / 555K GPUs / 2 GW by 2026 (all 0-0 to 0-1).
- DeepSeek/High-Flyer's GPU fleet (~10K H100 + 10K H800 + 30K H20 + 10K A100) and $1.63B capex (0-0).
- DeepSeek outages and halted signups as a serving-capacity signal (0-0).
- OpenAI/Oracle deploying 64K GB200 at Abilene by end-2026 (1-0).
The pattern: self-built and Chinese-lab capacity is the hardest to substantiate. Meta may well hold serious compute; the evidence set just could not confirm it.
09 · What to watchWhat to watch (the open questions that would change the ranking)
- Real serving metrics. The entire reliability axis is built on proxies. The first lab to publish honest tokens/sec, concurrent-user ceilings, and p99 under load changes this analysis.
- Meta's true buildout. If Prometheus (~1 GW) and Hyperion (up to 5 GW) are real and on schedule, Meta jumps into the top tier.
- xAI's 2026 reality. Colossus 1 is confirmed; everything after it is not. Satellite-imagery analysts already dispute the Colossus 2 "1 GW" framing (Tom's Hardware put on-site cooling at ~350 MW).
- The training-versus-inference split. None of the contracted capacity is broken out by how much is reserved for training versus available for serving, or, for Google TPU, how it is split between DeepMind and external tenants like Anthropic. That split is what actually determines service consistency.
Bottom line
For consistent, top-level service delivery specifically, the order is Google DeepMind and Anthropic at the top (one by vertical integration, the other by diversification), OpenAI third (biggest ceiling, tightest present), xAI fourth (strong cluster, thin and unverified forward capacity), and Meta, DeepSeek, and Mistral unranked for lack of verifiable data.
But hold the dominant caveat in front of all of it: this is a mid-2026 snapshot of a market where the announced numbers are roughly 20-to-1 against the online ones, and where even the best-positioned lab took a capacity-driven global outage last week. The leaderboard you actually care about is "online and serving," and on that board, almost everyone is smaller than their headlines.
Sources
Independent / third-party (strongest):
- Epoch AI, "OpenAI Stargate: Where the US Sites Stand" (satellite imagery, Apr 2026)
- SemiAnalysis, "TPUv7: Google Takes a Swing"
- SemiAnalysis, "xAI Colossus 2: First Gigawatt Datacenter"
- Tom's Hardware, "Colossus 2 nowhere near 1 GW" (satellite analysis)
- The National, "Claude hit by major global outage due to unexpected capacity constraints" (Jun 2, 2026)
- Scientific American, "What is the AI compute crunch?"
First-party (directionally reliable, promotional, unaudited):
- OpenAI, "Stargate advances with Oracle partnership"
- Anthropic, "Anthropic and Amazon compute"
- Anthropic, "Expanding our use of Google Cloud TPUs"
- Anthropic, "Google + Broadcom partnership compute"
- Google Cloud, TPU v7x documentation
Secondary / contested (used with caution):
- CSIS, "DeepSeek, Huawei, Export Controls"
- Data Center Frontier, "Ownership and power challenges in Meta's Hyperion and Prometheus"
- The Decoder, "The AI industry is running out of compute"
Research method: deep-research harness, 6 search angles, 25 sources fetched, 113 claims extracted, top 25 verified via 3-vote adversarial verification (17 confirmed, 8 killed). 108 agents, ~3.4M tokens. GW-to-fleet reconciliation applied via the sizing-data-centers skill.