For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for CoreWeave, Inc. (CRWV) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-09-21): Designation Low · Gem Score -36 (−100…+100 Quality+Value blend) · Quality -81 · Value 0 · Sentiment 36 (timing only, not weighted)
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
CoreWeave, Inc.
CRWV NASDAQCoreWeave, Inc. is an American artificial intelligence cloud computing company that provides high-performance infrastructure tailored to modern AI and accelerated computing workloads. Headquartered in Livingston, New Jersey, CoreWeave focuses on delivering large-scale GPU-based compute resources for applications such as generative AI, machine learning, visual effects rendering, and other data- and compute-intensive tasks. Its platform combines specialized hardware, networking, and storage with orchestration and chip management software to help enterprises and AI developers deploy, scale, and optimize complex models efficiently. CoreWeave positions itself as a purpose-built cloud for AI, offering flexible, high-throughput infrastructure and technical support aimed at organizations that require more specialized performance than general-purpose cloud providers typically offer. The company serves a global client base that includes AI labs, fast-growing startups, and large enterprises across technology, media, and other data-driven industries.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): -2.81
Total Equity: $3.34B
Shares: 436,000,000
Total Debt: $21.37B
Cash: $3.13B
EBITDA: $2.41B
Total Debt: $21.37B
Cash: $3.13B
Revenue: $5.13B
Shares: 436,000,000
Revenue: $5.13B
Revenue: $5.13B
Revenue: $5.13B
Total Equity: $3.34B
Tax Rate: 4.0%
Equity: $3.34B
Total Debt: $21.37B
Cash: $3.13B
Current Liabilities: $16.44B
Long-Term Debt: $14.67B
Total Debt: $21.37B
Total Equity: $3.34B
Shares: 436,000,000
Shares: 436,000,000
CapEx: -$10.31B
Shares: 436,000,000
Stock Price: $87.85
Net Income: -$1.17B
Industry Benchmarks
Income Statement (Annual)
Last updated: Aug 24, 2026 5:30am (28d ago)| Metric | 2023 | 2024 | 2025 |
|---|---|---|---|
| Revenue | $229.0M | $1.9B | $5.1B |
| Cost of Revenue | $69.0M | $493.0M | $1.5B |
| Gross Profit | $160.0M | $1.4B | $3.7B |
| Operating Expenses | $174.0M | $1.1B | $3.7B |
| Operating Income | -$14.0M | $324.0M | -$46.0M |
| Net Income | -$594.0M | -$863.0M | -$1.2B |
| EBITDA | $89.0M | $1.2B | $2.4B |
| EPS | $-3.09 | $-4.30 | $-2.75 |
| EPS (Diluted) | $-3.09 | $-4.30 | $-2.81 |
Balance Sheet (Annual)
Last updated: Aug 24, 2026 3:30am (28d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Cash & Equivalents | — | $217.0M | $1.4B | $3.1B |
| Total Current Assets | — | — | $1.9B | $7.5B |
| Total Assets | — | — | $17.8B | $49.3B |
| Current Liabilities | — | — | $5.0B | $16.4B |
| Long-Term Debt | — | — | $5.5B | $14.7B |
| Total Liabilities | — | — | $16.5B | $46.0B |
| Total Equity | -$11.0M | -$597.0M | $1.3B | $3.3B |
| Retained Earnings | — | — | -$1.5B | -$2.6B |
Cash Flow (Annual)
Last updated: Aug 24, 2026 5:30am (28d ago)| Metric | 2023 | 2024 | 2025 |
|---|---|---|---|
| Operating Cash Flow | $1.8B | $2.7B | $3.1B |
| Capital Expenditure | -$2.9B | -$8.7B | -$10.3B |
| Free Cash Flow | -$1.1B | -$6.0B | -$7.3B |
| Acquisitions (net) | $0 | $0 | -$108.0M |
| Net Debt Issued / (Repaid) | $1.4B | $6.4B | $8.4B |
| Dividends Paid | — | — | — |
| Stock Buybacks | — | — | — |
| Net Change in Cash | $473.0M | $1.6B | $2.1B |
Growth Trends (YoY %)
Last updated: Aug 24, 2026 5:30am (28d ago)| Metric | 2024 | 2025 |
|---|---|---|
| Revenue Growth | +736.2% | +167.9% |
| Gross Profit Growth | +788.8% | +158.6% |
| Operating Income Growth | +2,414.3% | -114.2% |
| Net Income Growth | -45.3% | -35.2% |
| EBITDA Growth | +1,233.7% | +102.9% |
Deep Analysis
Risk : Reward — upside vs downside from this company's own quarters
Computed 2026-09-18 02:42A +1σ run of quarters pays -186%; a −1σ run costs 100%. Ratio -1.9:1 (μ 152.7%, σ 98.5% , 6 pairs).
Older method (repeat-worst-quarter): 2.9 : 1
| Case | Growth | Margin | Fair value | vs price ($87.85) |
|---|---|---|---|---|
| Bull — recovery | +150% | 9.2% | $340.91 | +288% |
| Base — stabilizes | +100% | 8.0% | $108.36 | +23% |
| Bear — keeps slipping | +50% | 6.8% | $23.84 | -73% |
| Stress — last quarter repeats | +134% | -25.4% | $0.00 | -100% |
| Upside — a +1σ run of quarters (v2) | +50% | -12.7% | $-75.45 | -186% |
| Stress — a −1σ run of quarters (v2) | +50% | -25.4% | $0.00 | -100% |
Narrative Economics
market-narrative step).
Growth Outlook
Analyzed 2026-08-24 05:39The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The raw trajectory is genuinely arresting: quarterly revenue went from $584M (Q3'24) to $2.58B (Q2'26) — a 4.4x lift in seven quarters, with sequential growth of $2.08B→$2.58B (+24% QoQ) as recently as last quarter. That is not decelerating in absolute terms; the "decelerating quarterly trend" flag is technically true on a percentage basis but misleading given the base effect. Gross margin at 71.7% is real and structurally interesting for an infrastructure business. But everything below the gross line is a horror show once you look honestly: 2025 operating income of -$46M on $5.1B revenue means SG&A + D&A + interest is eating the entire gross profit stack, and 2025 FCF of -$7.25B against $10.3B capex tells you the "revenue" is being manufactured by borrowing to buy Nvidia GPUs and renting them back to a handful of hyperscale-adjacent customers. The $21.4B debt against $3.3B equity (D/E 6.4x) and current ratio of 0.46 is a genuine solvency configuration, not a growth-company balance sheet.
The synthesis "Priced for Perfection" verdict and the Market Forces "Headwinds" call are directionally right but I think they still understate the debt-cycle risk. GPUs depreciate on a 3-5 year useful life; H100s are already being repriced downward as B200/GB200 supply ramps. CoreWeave's $10B annual capex isn't optional maintenance — it's a treadmill. If contracted revenue (Microsoft is reportedly ~60%+ of revenue) renegotiates on renewal at 20-30% lower pricing to reflect commoditized GPU-hours, the unit economics invert violently because the debt service is fixed. The thesis evaluation's -11 score feels too generous; the bull weights assume "AI demand exceeds forecasts" is independent of "hyperscalers own this market by 2027," but they aren't independent — the same hyperscalers that are CRWV's largest customers are building competing capacity. That's the contradiction the models danced around: Microsoft is simultaneously CoreWeave's lifeline and its most credible executioner.
The contrarian bull case worth stating: at $87.85 and $48B market cap on a $2.58B quarterly run-rate (~$10.3B annualized), the forward EV/revenue is roughly 6.7x including net debt of $18B — not obviously insane if you believe 2027 revenue lands at $18-22B and gross margins hold. The narrative-mode assessment (90% story) is correct but incomplete: the story has a hard cash test in 2026-27 when the first tranches of that $21B debt need refinancing or extending, and interest coverage will be the metric that matters, not revenue growth. If rates stay elevated and the AI capex cycle even mildly disappoints, CRWV's equity is a call option on refinancing. The insider activity — pure option-exercise-and-sell mechanics on a single day (Aug 20), no discretionary open-market buying — is what you'd expect but tells you insiders aren't stepping up here either.
I agree with the synthesis verdict but land more bearish than "priced for perfection." Perfection is at least a coherent scenario; my read is that the model is structurally fragile because customer concentration + debt load + GPU depreciation curve create a scenario where even good execution can produce equity impairment. Fair value on a probability-weighted basis is closer to $40-55 (haircut for refinancing risk, customer concentration renegotiation, and terminal margin compression to 15-20% EBITDA rather than the 30%+ the bull case needs). The gap to $87.85 is the narrative premium. I'd short-bias this into strength, but sizing must respect that AI-infra momentum can push it to $120+ before reality asserts — this is not a name to be short-heavy on before a catalyst. The catalyst to watch: Microsoft contract renewal terms, any Nvidia allocation shift toward hyperscalers-direct, or the first covenant-adjacent quarter where operating cash flow can't cover interest + maintenance capex.
GPT Reading
CoreWeave’s income statement is screaming “real demand,” but the balance sheet and cash flow statement are screaming something louder: this is a financing structure masquerading as a software multiple. Revenue went from $229M in 2023 to $1.92B in 2024 to $5.13B in 2025, and the quarterly run-rate has kept climbing to $2.58B in the June 2026 quarter, up from $1.21B a year earlier. On an annualized basis that is already a $10B+ revenue pace. Gross profit scaled too, with 2025 gross profit at $3.68B on $5.13B of revenue, a very strong 71.7% gross margin for what is fundamentally compute infrastructure. That combination is what keeps the stock afloat at $87.85 and a $48.45B market cap: if the revenue is durable and the gross margin is real, investors can imagine a future where fixed costs get absorbed and earnings inflect fast.
But the rest of the numbers don’t support paying up as if that future is close. In 2025, operating income was still negative $46M despite the huge gross profit pool, and net loss was $1.17B. More importantly, operating cash flow of $3.06B is being overwhelmed by capex of $10.31B, leaving free cash flow at negative $7.25B. That’s not a temporary rounding error; it is the business model at this stage. The latest annual balance sheet shows $21.37B of debt against just $3.13B of cash and $3.34B of equity, with a current ratio of 0.46 and debt-to-equity above 6x. This is not a company merely “investing ahead of growth”; it is one that must continuously win both customer demand and capital market confidence. At $48.45B equity value, investors are effectively underwriting years of heavy external financing while assuming that the assets being built will earn very high utilization and attractive pricing in an industry where supply responses are coming from the deepest-pocketed competitors in the world.
The quarter-to-quarter pattern also argues against treating this like a clean software compounding story. Revenue is still growing rapidly, but the net margin remains deeply negative and erratic: -32.1%, -23.9%, -8.1%, -28.7%, -35.6%, then -24.3% in the most recent quarter. That is not what an imminent operating leverage story usually looks like. Even on the favorable metric, valuation is rich. If I annualize the latest quarter’s $2.58B, the stock is around 4.7x current run-rate sales; on trailing 2025 revenue it is about 9.4x sales. For a capital-light software name, maybe you can stretch there. For a company with negative free cash flow of more than $7B, leverage over $21B, and an asset base that will require relentless refresh in GPUs and data-center infrastructure, I can’t. The market is valuing CoreWeave as if today’s extraordinary AI demand, pricing, and customer urgency will persist long enough for the company to delever into strength. That is too generous.
The best case against my view is straightforward: the scale is arriving so quickly that the valuation could compress naturally even if the stock goes nowhere. A business that was at $982M of revenue in Q1 2025 and is at $2.58B by Q2 2026 is not decelerating into irrelevance; it is still compounding at a pace few public companies ever achieve. If gross margins can remain around 70% while SG&A and other operating costs grow slower than revenue, CoreWeave may be much closer to meaningful operating profit than GAAP net income suggests. The operating income swing from positive $324M in 2024 to negative $46M in 2025 despite much larger revenue hints that accounting noise, financing costs, and depreciation are muddying the underlying earnings power. If capex is front-loaded and this installed capacity monetizes over several years, today’s negative free cash flow could look like a deliberate land-grab rather than structural value destruction. In that world, 2026 or 2027 revenue could plausibly exceed $10B-$12B, making the current market cap less extreme than it appears on trailing numbers.
What would change my mind is evidence that CoreWeave can translate hypergrowth into self-funding economics before the balance sheet becomes the story. Specifically, I would want to see at least two consecutive quarters of operating margin clearly positive, net losses narrowing toward better than -10% margin, and annualized operating cash flow rising enough that free cash flow losses shrink materially from the 2025 level even with continued capex. I would also need to see debt growth slow relative to revenue growth and liquidity improve from the current sub-0.5 current ratio. If the company can exit a year with a revenue run-rate above $12B, gross margin still north of 65%, and free cash flow burn cut by half without a major equity raise, the bull case gets much more credible. Until then, this looks like a phenomenal business franchise possibility attached to a valuation that assumes near-perfect execution in a brutally capital-intensive market.
Grok Reading
The numbers on CoreWeave describe a company sprinting up a revenue cliff while loading the balance sheet to the breaking point. Revenue scaled from $229M in 2023 to $1.92B in 2024 to $5.13B in 2025, and the two most recent quarters—$2.08B and $2.58B—imply a run-rate already approaching $10B if the pace holds. Gross margin of 71.7% is genuinely elite for infrastructure and shows the GPU rental model can price like software when capacity is scarce. That is the entire bull case in the raw data: hypergrowth plus software-like gross profit on a purpose-built AI stack. Everything else is a red flag. Net losses are not shrinking in absolute terms; the last two quarters alone lost $740M and $626M. Free cash flow was −$7.25B on $10.31B of capex against only $3.06B of operating cash flow. The balance sheet carries $21.37B of debt against $3.13B of cash and $3.34B of equity, a 6.4× debt-to-equity ratio and a 0.46 current ratio that leaves almost no margin for a demand hiccup or a refinancing freeze. At $87.85 and a $48.5B market cap the stock trades at roughly 7.5× trailing sales and 13× EV/revenue while still deeply unprofitable—pricing that assumes the growth curve continues and the capital intensity magically normalizes.
The trajectory of losses and leverage is the contradiction the models underweight when they simply label this “pre-profit growth.” Operating margin is still negative (−0.9% annually), ROE is −35%, and ROIC is essentially zero. Sequential revenue is still rising, but the year-over-year growth rate has already decelerated from the 700%+ phase into the high-100s, and the secondary signals correctly flag low revenue confidence and a decelerating quarterly trend. Insider activity in August 2026 was a cluster of option exercises paired with open-market sales—not catastrophic, but consistent with liquidity-taking rather than accumulation at these levels. The company is generating real operating cash, which is better than pure cash-combustion startups, yet that cash is being immediately overwhelmed by capacity build-out that has not yet proven it can earn its cost of capital once GPU scarcity eases.
The strongest case against this read is straightforward and must be taken seriously: if AI training and inference demand remains supply-constrained for another two to three years, CoreWeave’s installed base becomes a toll bridge with 70%+ gross margins and the current −$7B FCF is simply growth capex that converts to owner earnings once utilization stabilizes. A $10B-plus revenue run-rate already in sight, combined with any sustained improvement in operating leverage, could make today’s 13× EV/sales look cheap in hindsight if the firm reaches $20B-plus revenue with mid-20s free-cash-flow margins. Customer lock-in via specialized clusters and multi-year contracts could blunt hyperscaler pricing pressure longer than the base rate suggests, and the positive operating cash flow demonstrates the unit economics work before growth spending. That path is coherent; it is also the path already fully discounted in the $48B equity value and the $21B debt stack.
I would reverse to a constructive stance only on clear evidence that the capital intensity is cresting—specifically, two consecutive quarters of free-cash-flow improvement toward breakeven, a material reduction in net debt or a lengthening of the maturity wall, and sequential revenue growth that holds above 15% without gross-margin compression below 65%. Absent those, the stock is a leveraged call option on perpetual GPU scarcity priced as if the option is already deep in the money.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
CoreWeave is scaling revenue explosively ($229M to $1.92B to $5.13B across 2023-2025, with gross margins in the low-70s) but the business consumes cash at an extraordinary rate: FCF was -$7.25B in 2025 on top of -$5.95B in 2024, and net income remains deeply negative (-$1.17B in 2025). Operating margin actually reversed from +16.9% in 2024 to -0.9% in 2025, undermining the story that scale brings profitability. OCF/NI of -2.96x and accruals at -14.4% of assets reflect the capex-and-financing-heavy model rather than clean earnings power. The balance sheet is the dominant fact: net debt of $18.21B against $3.16B liquid cash, short-term debt of $6.71B alone exceeding cash, and a computed runway under two quarters at the current burn. Altman Z of 0.44 sits squarely in the distress zone. Meanwhile diluted share count doubled from 218M to 436M in a single year (a 50.7% CAGR) and SBC runs 12.3% of revenue with zero buyback offset, so per-share economics are being crushed even as top-line compounds. Insider behavior compounds the picture: 83 sales and zero open-market buys over the last 12 months ($94.8M sold), and the most recent tape shows the CEO and multiple officers systematically exercising options and selling into the market. Nothing here suggests fraud, but this is a capital-hungry, dilution-heavy, debt-laden business whose survival depends on continued access to financing and customer demand holding up.
Verify before trusting this (6)
- Debt maturity ladder and covenants on the $6.71B short-term debt, and whether customer prepayments or delivery-linked financings are collateralized against GPU assets
- Customer concentration (Microsoft and other hyperscaler share of revenue) and contract duration/take-or-pay structure
- Capex commitments and lease obligations relative to booked revenue backlog
- Terms of any convertible notes and warrant overhang contributing to the diluted-share explosion
- Whether the 2025 OpM decline reflects one-time IPO-related SBC or structural cost creep
- 10b5-1 plan status behind the August 2026 insider sales cluster
None surfaced.
None surfaced.
CRWV sits at the epicenter of the AI-infrastructure narrative, which is currently one of the most intense story arcs in the market. The tape is mildly risk-on, the VIX is subdued, and the archetype (platform-monopoly, purpose-built GPU cloud) is exactly what momentum money wants to own. Recent news flow is overwhelmingly narrative-reinforcing: Tepper piled in, an HRT/Vera Rubin deal amplifies the specialized-AI-demand story, Shkreli covered his short, and pieces are running with headlines like 'Stock Will Double.' For a high-beta, story-driven, pre-profit name, that is the sentiment cocktail that keeps a bid under the stock. Counter-pressure is real but secondary: BCA Research's capital-destruction warning, an insider sale by the CEO, and a bear framing around debt-fueled capex are starting to seed doubt about durability. The stock is ~90% story per the brief, so any crack in the neocloud narrative would land hard here specifically because there is no fundamental floor to catch it. For now the narrative is intact and intensifying, but durability is only 'moderate' and the recent 167.9% is decelerating vs the 373.4% long-term rip, hinting the mania is maturing rather than accelerating. Net: still a tailwind, but not the strong-tailwind reading it would have been a quarter ago.
Verify before trusting this (5)
- Whether the BCA-style 'debt-funded capex destroys capital' frame gains traction with sell-side or other macro strategists
- Any hyperscaler pricing action (AWS/Azure GPU cuts) that would crack the pricing-power leg of the bull story
- Follow-on insider sales or secondary offerings that would signal the story is being monetized from the inside
- NVDA-linked catalysts (Vera Rubin ramp, allocation news) that either reinforce or undermine the specialized-AI moat narrative
- A meaningful VIX spike or rates move that would disproportionately hit high-beta pre-profit names
The world is re-plumbing compute: training and, increasingly, inference workloads require dense, liquid-cooled, high-interconnect clusters that legacy general-purpose cloud footprints were not designed for. That mismatch is CoreWeave's entire reason to exist and it is a multi-year build, not a quarter. Against that, the financing environment is the counterweight — a 4.69% 10-year, a barely positive curve and macro headwinds make debt-funded GPU fleets more expensive exactly when the asset's useful economic life is uncertain (silicon generations turn faster than depreciation schedules). The honest structural read: demand for AI compute is very likely to keep growing for years, but the ECONOMICS of being the leveraged landlord of rapidly-obsoleting silicon are far less certain than the demand curve. Growth in revenue is the high-confidence part; growth in durable earnings power is the contested part.
When we made this prediction on Aug 24, 2026, CRWV was $86.38. We expect it to be $76.00 by Feb 2027. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 24, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.