Deep Analysis of CoreWeave (CRWV)#

Sector: Technology

Industry: Software — Infrastructure (specialized AI cloud / “neocloud” / GPU-as-a-Service)

This article is for informational purposes only and is not investment advice. Figures were gathered from public sources listed at the end.

Introduction#

CoreWeave, Inc. is an AI-focused cloud infrastructure company that rents high-performance NVIDIA GPU compute, storage, networking, and orchestration software to AI labs, hyperscalers, and enterprises. Founded in 2017 (originally as an Ethereum cryptocurrency miner) and headquartered in Livingston, New Jersey, the company pivoted to GPU cloud services and completed a high-profile Nasdaq IPO in March 2025. Branding itself “The Essential Cloud for AI,” CoreWeave positions itself between the AI models and the silicon — operating purpose-built “AI factories” optimized for large-scale training and, increasingly, inference. It has become the largest of the new wave of “neocloud” providers and the fastest cloud platform in history to surpass $5 billion in annual revenue, serving a customer roster that now includes Microsoft, OpenAI, Meta, Anthropic, Google, and NVIDIA itself.

Fundamental Analysis#

CoreWeave is a hyper-growth company with extraordinary top-line momentum but a financially fragile balance sheet. Revenue is expanding triple digits and contracted demand is enormous, yet the business remains deeply unprofitable on a GAAP basis and burns cash heavily because it must buy GPUs and build data centers years ahead of recognizing the revenue. The defining tension is leverage: CoreWeave funds its build-out largely with debt, and interest expense now consumes nearly all of its operating profit. In short, the company is operationally strong (demand, margins, scale) but financially stretched (debt, cash burn, refinancing dependency). Whether it is “weak” or “strong” depends almost entirely on continued access to cheap capital and on AI demand staying robust.

• Revenue: Q1 FY26 (ended March 31, 2026) revenue was approximately $2.08 billion, up ~112% year-over-year. FY2025 revenue was roughly $5.1 billion; management guides FY2026 revenue to ~$12–13 billion.
• Profitability: Q1 FY26 GAAP net loss was ~$740 million (net loss margin ~-36%), widening from a ~$315 million loss a year earlier. FY2025 net loss was ~$1.2 billion.
• Margins: Gross margin is high (~69–74%), and adjusted EBITDA margin is strong (~56%, ~$1.16 billion in Q1). However, after ~$536 million of net interest expense, adjusted operating income was only ~$21 million — illustrating how debt service erodes profitability.
• Backlog: Revenue backlog reached roughly $99.4 billion as of Q1 FY26 (up from ~$66.8 billion at end-2025), providing multi-year revenue visibility (management has noted roughly 75% is weighted to within four years).
• Capacity: Surpassed 1 GW of active power; ~3.5 GW of total contracted power; targeting ~1.7 GW online by year-end 2026 and more than 8 GW by 2030.
• Capital intensity: 2026 capex guidance is ~$31–35 billion. Q1 free cash flow was roughly -$4.7 billion; trailing levered free cash flow is sharply negative.
• Leverage: Total debt is large (roughly $25–30 billion depending on source and date), with a debt-to-equity ratio reported near 700%+ and a current ratio around 0.3. Recent unsecured notes priced at ~9.75%, while newer asset-backed facilities have achieved investment-grade ratings, signaling a bifurcated and improving-but-still-expensive cost of capital.
• Valuation: Market cap ~$63 billion; price-to-sales ~9.5x; no meaningful P/E given losses. Shares traded ~$115–119 in mid-June 2026, within a 52-week range of ~$63.80–$187.00.

Key Products or Services#

CoreWeave’s offering is a full-stack, vertically optimized platform for running AI workloads on NVIDIA hardware, spanning raw compute, orchestration software, storage, developer tools, and managed services. Revenue is overwhelmingly driven by committed, multi-year GPU compute contracts with large AI labs and hyperscalers, increasingly supplemented by higher-value software and managed-inference services. The platform’s differentiation lies less in the chips themselves (which competitors can also buy) and more in CoreWeave’s software, networking, and operational tuning that extract near-bare-metal performance at scale.

• CoreWeave Cloud Platform: Large-scale GPU and CPU compute delivered on usage-based and committed contracts — the core revenue engine.
• CoreWeave Kubernetes Service (CKS) & SUNK: Kubernetes-native infrastructure control and a Slurm-on-Kubernetes scheduler that lets customers move fluidly between training and inference so GPUs rarely sit idle.
• Mission Control: Node, rack, and fleet lifecycle management for reliability and uptime at hyperscale.
• Weights & Biases (acquired): A widely used AI developer platform for experiment tracking, model evaluation, and now agent/sandbox tooling — a strategic move up the value stack toward developers.
• Storage & data movement: AI Object Storage and the Local Object Transport Accelerator (LOTA), purpose-built for high-throughput AI workloads.
• Runtime acceleration & developer tools: Tensorizer and the acquired open-source notebook Marimo, extending the developer toolchain.
• CoreWeave Federal: A newer offering extending the platform to U.S. government and public-sector use cases.
• CoreWeave Ventures: A venture fund investing in AI startups (which can become future customers).

Moats, Strengths and Weaknesses#

Moats#

• Privileged NVIDIA relationship: CoreWeave is repeatedly first-to-market with the newest NVIDIA platforms — it was the first AI cloud to validate Vera Rubin NVL72 at rack scale, posts record MLPerf benchmarks, and was named an NVIDIA Exemplar Cloud. NVIDIA holds an equity stake and has agreed to backstop unsold capacity, creating a deep, mutually reinforcing tie.
• Performance reputation: It is the only provider to earn SemiAnalysis’s top Platinum ClusterMAX rating across consecutive rankings, a credible third-party signal of engineering quality that is hard to replicate.
• Scale and capital-markets access: As the largest neocloud, it has pioneered GPU-/HPC-backed financing as an emerging asset class, including the first investment-grade-rated GPU-backed facility — a funding moat smaller rivals cannot easily match.
• Contracted backlog and switching costs: A ~$99 billion backlog and ten customers committed to $1 billion+ create stickiness and visibility.

Strengths#

• Triple-digit revenue growth and high gross margins (~70%).
• A blue-chip customer base spanning the four largest AI model developers (OpenAI, Anthropic, Google, Meta) plus Microsoft and NVIDIA.
• Software differentiation (SUNK, Mission Control, Weights & Biases) that elevates it above a commodity GPU-rental business.
• Improving credit profile on asset-backed debt and growing institutional acceptance of its financing model.

Weaknesses#

• Heavy debt and interest burden: Roughly $25–30 billion of debt with interest expense that nearly wipes out operating income; the model is acutely sensitive to interest rates and credit-spread widening.
• Persistent losses and cash burn: Large GAAP net losses and deeply negative free cash flow; the company depends on continuously open capital markets.
• Customer concentration: Microsoft was ~62% of 2024 revenue and ~67% of 2025 revenue; the top handful of customers still dominate, so losing one would be material (though diversification is improving).
• Single-supplier dependency: Near-total reliance on NVIDIA hardware exposes it to supply, pricing, and roadmap risk.
• Leased footprint / failed verticalization: CoreWeave leases most of its data centers and remains a large tenant; its attempt to acquire data-center owner Core Scientific (to verticalize) was rejected by Core Scientific shareholders in October 2025, leaving roughly $10 billion of long-term lease obligations and less control over power and real estate.
• Litigation overhang: Multiple securities class-action lawsuits allege the company misled investors about scaling and data-center delivery timelines.
• Customers are also competitors: Hyperscalers (Microsoft, Meta, Google) build their own capacity even as they rent from CoreWeave.

News, Events and Partnerships#

The last six months have been dominated by a wave of mega-contracts, aggressive debt financing, and index-inclusion catalysts — broadly positive for revenue visibility but reinforcing the leverage narrative. The strategic headline is that CoreWeave has now signed all four of the largest AI model developers, sharply reducing the single-customer risk that overshadowed its IPO. Counterbalancing this, the company continues to raise expensive debt and faces ongoing AI-bubble scrutiny.

• Meta — ~$21 billion expansion (April 2026): A new agreement through December 2032 (on top of a prior ~$14.2 billion arrangement), including early deployments of NVIDIA’s Vera Rubin platform — a strong diversification win. (Positive.)
• Anthropic (Q1 2026): A multi-year agreement to run Claude workloads at production scale, completing CoreWeave’s set of the four leading AI labs. (Positive.)
• OpenAI (cumulative ~$22.4 billion): Expanded across 2025 into one of CoreWeave’s anchor relationships. (Positive, but tied to OpenAI’s own funding/profitability.)
• NVIDIA (January 2026): Expanded collaboration to build 5+ GW of AI factories by 2030; NVIDIA invested $2 billion at $87.20/share and CoreWeave committed to adopting more NVIDIA products. (Positive — but feeds “circular financing” criticism.)
• Jane Street (April 2026): ~$6 billion cloud commitment plus a $1 billion equity investment at $109/share. (Positive.)
• Financing blitz: Closed an $8.5 billion investment-grade GPU-backed facility (March), a $3.1 billion syndicated HPC-backed facility (May), priced $2.75 billion of 9.75% senior notes (April), and announced additional senior notes in June. (Mixed — validates the financing model but underscores rising leverage and high coupons.)
• Technology milestones: Record MLPerf Training v6.0 results (including a ~2-minute DeepSeek-V3 run), first to validate Vera Rubin NVL72, and new “agentic AI” capabilities. (Positive for competitive positioning.)
• Nasdaq-100 inclusion (effective June 22, 2026): Expected to drive passive index inflows. (Positive technical catalyst.)
• CrowdStrike partnership: Collaboration on a secure AI cloud foundation for the agentic era. (Positive.)
• Core Scientific deal terminated (October 2025): Shareholders rejected the ~$9 billion all-stock acquisition; the two remain commercial partners. (Negative for verticalization strategy.)

Government Integration#

CoreWeave’s government footprint is early-stage but strategically deliberate, primarily via its new public-sector arm and a high-profile federal science initiative rather than disclosed large dollar contracts. There is no publicly disclosed evidence of CoreWeave being a recipient of major federal grants or multi-billion-dollar government procurement contracts to date; the opportunity is positioned as forward-looking, leveraging national interest in domestic AI and energy infrastructure.

• CoreWeave Federal: Launched to extend the platform to U.S. government and public-sector AI workloads, positioning the company for future federal demand.
• DOE Genesis Mission: CoreWeave joined this U.S. Department of Energy initiative focused on accelerating discovery science, strengthening national security, and advancing U.S. energy innovation — aligning it with federal AI/compute priorities.
• Policy tailwinds: Broad U.S. policy emphasis on domestic AI leadership, data-center build-out, and energy supply indirectly benefits CoreWeave, though it also exposes the firm to NVIDIA export-control dynamics and permitting/power constraints.

Social Sentiment#

Retail and social sentiment is currently net bullish but highly volatile and event-driven. Reddit chatter has been steadily positive (sentiment readings in the mid-60s to low-70s on a 0–100 scale), buoyed by the Nasdaq-100 inclusion and the parade of mega-deals, and options flow has skewed aggressively bullish (e.g., heavy June-expiry call buying with tens of millions in premium on single strikes). At the same time, CoreWeave is widely treated as a “story stock” and a leveraged proxy for the broader AI-infrastructure trade — it swung from roughly $91 to $132 within weeks in mid-2026 — and bearish voices repeatedly cite the debt load, cash burn, insider selling, and “AI bubble” fears, with short interest having spiked to around 10% during the late-2025 drawdown. Sentiment thus oscillates sharply with each macro AI-capex headline.

Insider Activity#

Insider activity is overwhelmingly and consistently one-directional: selling. Over the trailing six months, public-market insider transactions were essentially all sales — reportedly on the order of 1,800+ disposals and effectively zero open-market purchases. Major sellers include early backer Magnetar Financial (well over $1.5 billion), co-founder/Chief Strategy Officer Brian Venturo ($885 million), co-founder/Chief Development Officer Brannin McBee ($365 million), CEO Michael Intrator ($361 million), and other insiders, with cumulative insider/early-investor sales exceeding $2.3 billion since the IPO; the CFO has also trimmed holdings. While much of this reflects post-lockup diversification and pre-arranged plans rather than necessarily a bearish signal, the complete absence of insider buying is a notable negative for sentiment-watchers.

Politician Activity#

Congressional trading in CRWV has been negligible and slightly net-negative. Public trackers show only a single congressional name associated with the stock over the past two years — Representative Cleo Fields — with small purchases (estimated ~$16,000) and small sales (estimated up to ~$33,000–$50,000, most recently around December 11, 2025), netting to a small negative figure. In short, CRWV is not a notable name in politician portfolios, and there is no meaningful bullish or bearish political-trading signal.

Institutional Activity#

Institutional positioning is genuinely mixed, reflecting the bull/bear divide over the stock. In the most recent reporting quarter, roughly 505 institutions added to positions while about 413 trimmed — a near-even split. NVIDIA and Jane Street are strategic equity holders, signaling deep partner conviction, and the Nasdaq-100 inclusion will mechanically pull in passive/index funds. On the other side, some large active managers reduced exposure. Analyst coverage skews constructive overall.

• Bullish signals: NVIDIA’s ~$2 billion equity stake and Jane Street’s $1 billion investment; ~23 of ~36 analysts at buy/strong buy; median price target ~$135 (with bullish outliers such as BNP Paribas at $192); index-inclusion inflows beginning June 22, 2026; constructive notes from Wells Fargo and Cantor Fitzgerald.
• Bearish/cautionary signals: Fidelity (FMR LLC) cut its position by ~59% (~4.7 million shares, $362 million) in Q1 2026; early backer Magnetar a persistent heavy seller; some bearish analyst views (e.g., Seeking Alpha “Strong Sell” arguments) flag junk-level unsecured coupons, lease liabilities, margin compression, and structural pricing risk; wide dispersion in price targets ($36 to ~$303) underscores genuine disagreement.

Political Landscape#

The macro and geopolitical backdrop is a double-edged sword: an unprecedented AI capital-expenditure supercycle is fueling CoreWeave’s demand, but the same environment carries bubble risk, rate sensitivity, and supply-chain/policy exposure. Hyperscaler capex is at historic highs (Meta alone guided to ~$115–135 billion for 2026; the four largest U.S. hyperscalers spent on the order of ~$400 billion in 2025), and the “world is structurally short compute,” supporting CoreWeave’s order book. But debt-funded, capital-intensive AI infrastructure is precisely the kind of business that suffers most if the cycle turns or financing tightens.

• AI capex supercycle (tailwind): Insatiable demand for training and inference compute, plus initiatives like the ~$500 billion Stargate program, underpin CoreWeave’s backlog.
• Power and land constraints (mixed): Electricity availability and data-center permitting are now the binding bottleneck for AI build-outs — favorable for those who lock in power, but a cost and timeline risk.
• Interest-rate and credit-spread sensitivity (headwind): As a heavily levered borrower, CoreWeave’s economics hinge on continued access to low-cost capital; rising rates or widening spreads directly pressure the model.
• “AI bubble” and circular-financing scrutiny (headwind): Investors increasingly question whether enormous backlogs (e.g., industry RPOs in the hundreds of billions) are durable when key counterparties like OpenAI are not yet profitable, and whether NVIDIA-investing-in-its-own-customers inflates demand.
• NVIDIA export controls and supply chain (geopolitical headwind): U.S.–China chip-export policy and GPU supply timing affect CoreWeave’s deployment cadence and component costs.

The Competition#

Companies compared: Nebius Group (NBIS), IREN Limited (IREN), Oracle Corporation (ORCL)

Nebius Group (NBIS)#

Nebius is the purest publicly traded analog to CoreWeave — an NVIDIA-backed neocloud renting GPU compute to AI labs and enterprises — but it is built and funded very differently, with a software-rich, full-stack platform (a legacy of its roots in the former Yandex cloud) and a markedly cleaner balance sheet. It competes head-to-head for the same AI workloads and even the same marquee customer (it landed a large Meta compute deal of its own).

• Q1 reported group revenue ~$399 million (up ~684%), with core AI run-rate revenue around $1.9 billion — roughly a fifth of CoreWeave’s scale.
• Profitable in 2025 (net income ~$101.7 million) versus CoreWeave’s large loss; closed Q1 with ~$9.3 billion cash and far less leverage.
• NVIDIA invested ~$2 billion; secured a ~$27 billion Meta capacity deal; targeting 3+ GW of contracted power by end-2026.
• Stock dramatically outperformed in 2026 (up well over 100% YTD by various measures); trades at a richer price-to-sales multiple than CRWV, reflecting its cleaner profile.
• Also entering the Nasdaq-100, alongside CoreWeave.

IREN Limited (IREN)#

IREN is a vertically integrated neocloud that began as a Bitcoin miner and is converting its owned power, land, and data centers to AI cloud — the asset-heavy, self-owned counterpoint to CoreWeave’s lease-heavy model. It competes for AI-cloud customers and, crucially, owns the infrastructure CoreWeave typically rents, giving it cost and control advantages but exposing it to a still-transitioning revenue base and crypto-price volatility.

• Transitioning from Bitcoin mining to AI cloud; Q3 FY26 total revenue ~$144.8 million (AI Cloud ~$33.6 million and rising), with sizable non-cash impairments tied to retiring mining hardware.
• Targeting a 150,000-GPU fleet and roughly $3.4–3.7 billion of AI-cloud annualized run-rate revenue by end-2026, anchored by a large Microsoft contract.
• Owns power and land (e.g., Childress, Texas; Sweetwater development), enabling faster onboarding and better pricing leverage.
• Heavy single-customer concentration (Microsoft) and execution risk on the mining-to-AI pivot.

Oracle Corporation (ORCL)#

Oracle is the most formidable scaled competitor for the very largest AI training contracts. Its Oracle Cloud Infrastructure (OCI) directly courts the same AI labs (notably winning enormous OpenAI/Stargate-linked capacity) and brings hyperscaler balance-sheet heft, a global footprint, and a full enterprise software stack that pure neoclouds lack. It is a direct competitor for mega-deals even though it is far more diversified than CoreWeave.

• A mega-cap, diversified enterprise software and cloud company — orders of magnitude larger than CoreWeave by revenue and market cap.
• Disclosed a record remaining-performance-obligation backlog (reported around $523 billion) on the back of massive AI cloud commitments — which itself sparked AI-bubble and free-cash-flow debates.
• Carries rising debt and negative free cash flow tied to AI build-out, echoing (at far larger scale) the same capital-intensity dynamics as CoreWeave.
• Advantages: global multi-region data centers, enterprise relationships, and the ability to bundle database/applications with compute.

How CoreWeave stacks up#

CoreWeave is the scale leader and the purest, most aggressive bet on AI compute demand among dedicated providers. Its backlog (~$99 billion), customer roster (all four top AI labs plus Microsoft and NVIDIA), and technology leadership (first-to-newest-NVIDIA, sole Platinum ClusterMAX) are unmatched among the pure-play neoclouds. No dedicated peer can match its revenue scale or contracted demand visibility.

The trade-off is financial fragility relative to peers. Against Nebius, CoreWeave offers vastly more scale and backlog but carries far more debt, deeper losses, and negative cash flow — Nebius is smaller but profitable, cash-rich, and less geared. Against IREN, CoreWeave is bigger and more diversified across customers but does not own its infrastructure, whereas IREN’s owned power and data centers give it structural cost control that CoreWeave forfeited when the Core Scientific deal collapsed. Against Oracle, CoreWeave is nimble, specialized, and first-to-new-hardware, but Oracle brings a fortress balance sheet, global footprint, and enterprise bundling that CoreWeave cannot replicate.

The net picture: CoreWeave wins on growth, scale, and AI-native specialization; it loses on balance-sheet quality, capital cost, and infrastructure ownership. It is the highest-beta way to play the neocloud thesis.

Metric (approx., latest available)	CoreWeave (CRWV)	Nebius (NBIS)	IREN (IREN)	Oracle (ORCL)
Business model	Lease-heavy GPU neocloud	Software-rich pure-play neocloud	Owned-infrastructure neocloud (ex-miner)	Diversified hyperscaler (OCI)
Latest quarterly revenue	~$2.08B	~$0.40B (group)	~$0.14B (AI ramping)	Mega-cap (multi-$B)
Profitability	Large GAAP loss	Profitable in 2025	Loss (transition/impairments)	Profitable; AI capex pressuring FCF
Balance sheet	High debt (~$25–30B)	Cash-rich (~$9.3B), low gearing	Asset-heavy, moderate leverage	Large but rising debt
Backlog / RPO	~$99B	Multi-year (smaller)	~$3.4–3.7B ARR target	~$523B RPO
Infra ownership	Mostly leased	Mixed	Owns power/land/DCs	Owns global DCs
Key risk	Leverage / refinancing	Scale / competition	Concentration / pivot execution	AI capex / FCF

The Proxy#

Nebius Group (NBIS)#

Nebius is the best single proxy for CoreWeave because it expresses essentially the same thesis — NVIDIA-powered GPU cloud capacity sold to AI labs and enterprises amid a compute shortage — while removing much of CoreWeave’s most controversial characteristic: extreme leverage. For investors who want exposure to the neocloud trade but are wary of CoreWeave’s debt-funded, cash-burning model, Nebius offers a cleaner, profitable, cash-rich alternative that nonetheless moves with the same demand drivers, shares NVIDIA as a backer, competes for the same customers, and is entering the same index. It is also a direct competitor, which satisfies the “proxy can also be a competitor” condition.

• Same core driver: Revenue and stock performance track the AI-compute demand cycle, just like CRWV.
• Cleaner profile: Profitable in 2025, ~$9.3 billion cash, materially lower leverage — a “lower-risk” way to play the theme.
• Shared validation: NVIDIA equity backing and a large Meta capacity deal mirror CoreWeave’s strategic anchors.
• Differentiated stack: A more developer-oriented, AWS/Azure-like software console gives it a distinct competitive angle.
• Caveat: Smaller scale and a richer valuation multiple mean it offers less backlog visibility and arguably less upside if CoreWeave executes flawlessly — it is the more balanced risk/reward, not the higher-ceiling bet.

The Big Picture for CoreWeave#

CoreWeave sits at the center of the most important infrastructure build-out of this decade. Its current product offering — purpose-built, high-performance GPU clouds with sophisticated orchestration software — is tightly aligned with where the AI market is going: from a training-dominated phase toward a far larger, more durable inference economy, where the ability to keep expensive GPUs continuously utilized and to deploy the newest NVIDIA silicon first becomes a real competitive edge. The company has clearly demonstrated demand fit by signing all four leading AI labs plus Microsoft and NVIDIA, and by amassing a backlog approaching $100 billion. On product-market alignment alone, CoreWeave looks well-positioned for the next five to ten years.

The harder question is financial durability. CoreWeave’s model only works if two conditions hold: AI compute demand keeps compounding, and capital markets keep funding the GPU and data-center build-out at manageable cost. The company has shown it can innovate on the financing side — pioneering GPU-backed lending and achieving its first investment-grade asset-backed facility — but its unsecured debt still prices at junk-like yields near 9.75%, interest expense already consumes nearly all operating profit, and free cash flow is deeply negative. In a benign environment, this leverage is “cheap fuel” that accelerates growth; in a downturn, rising rates or a demand air-pocket could turn that same leverage against the company quickly. The failed Core Scientific acquisition also means CoreWeave continues to rent rather than own much of its physical footprint, leaving it with large lease obligations and less control over power.

There are three plausible futures. In the bull case, AI demand stays insatiable, CoreWeave converts its backlog into high-margin recurring revenue, achieves operating leverage as it scales, refinances debt at progressively lower rates, and grows into — then beyond — its valuation, cementing itself as a durable fourth pillar of cloud alongside the hyperscalers. In the base case, it remains a fast-growing but volatile, capital-hungry business whose stock swings violently with each AI-capex headline and financing event, delivering strong revenue growth but uneven profitability. In the bear case, the AI capital cycle cools, utilization or pricing disappoints, financing costs rise, and the leverage that powered the ascent becomes an existential constraint.

On balance, CoreWeave is more likely to be a beneficiary of where the market is heading than a victim of it — the demand signal is real, its NVIDIA alignment is a genuine advantage, and the shift toward inference plays to its strengths. The principal threat is not that the market leaves CoreWeave behind, but that CoreWeave’s balance sheet cannot survive a rough patch in an otherwise growing market. It is best understood as a high-conviction, high-volatility bet on the AI infrastructure thesis: substantial upside if demand and financing cooperate, with outsized downside if either falters. The insider selling and mixed institutional flows counsel caution on the equity even for those bullish on the underlying business.

Anyone using this analysis should weigh it against their own risk tolerance and consult a licensed financial professional before making investment decisions; nothing here is a recommendation to buy or sell.

Sources#

• Yahoo Finance — CRWV quote, key statistics, and analyst data (finance.yahoo.com/quote/CRWV)
• CNBC — CoreWeave Q1 FY26 earnings coverage; Meta $21B deal; Core Scientific vote (cnbc.com)
• CoreWeave Investor Relations & press releases — Q1 2026 results, deal announcements (investors.coreweave.com, coreweave.com/news)
• U.S. SEC EDGAR — CoreWeave Forms 8-K / 425 (earnings, Meta, NVIDIA, Jane Street, debt facilities, Core Scientific merger) and IREN Forms 8-K (sec.gov)
• Simply Wall St — CRWV future analysis and bull/bear summary (simplywall.st)
• CNN Markets — CRWV news, revenue growth, litigation headlines (cnn.com/markets/stocks/CRWV)
• Robinhood — real-time CRWV price and 52-week range (robinhood.com)
• Public.com — CRWV earnings detail (public.com)
• The Motley Fool — CoreWeave vs. Nebius; CoreWeave news analysis (fool.com)
• Yahoo Finance / GuruFocus — Nebius vs. CoreWeave comparison (finance.yahoo.com)
• UK Investor Magazine — Nebius vs. CoreWeave neocloud comparison (ukinvestormagazine.co.uk)
• Tech Fund / Tech Investments — NeoCloud economics (techinvestments.io)
• ABI Research — leading neocloud profiles (abiresearch.com)
• Northflank; RunPod; aimultiple; computeprices.com — neocloud/GPU provider comparisons
• TradingKey; MLQ.ai; SBO Financial — IREN/NBIS/CRWV neocloud analyses
• DatacenterDynamics — Meta–CoreWeave deal; Core Scientific deal termination (datacenterdynamics.com)
• Core Scientific Investor Relations & CNBC/Yahoo/Sherwood/Verdict/Alpha Spread/StockTitan — Core Scientific shareholder vote and termination
• Quiver Quantitative — congressional trades, insider trades, institutional 13F flows, social discussion summary (quiverquant.com)
• AltIndex; Capitol Trades — CRWV congressional trading statistics
• Kavout — CoreWeave CFO insider selling analysis
• MarketBeat — CRWV insider trading activity
• TipRanks — analyst ratings, Smart Score, Nasdaq-100 inclusion, options flow (tipranks.com)
• StocksToTrade — CRWV catalysts, leverage and FCF figures
• Barchart — BNP Paribas upgrade and consensus targets
• TradingView (Invezz) — bearish sentiment / short interest / AI-bubble context
• Augment Market Pulse — “four biggest AI labs” deal summary
• CoinCodex — supplementary price/sentiment context