Wondy.
KO|EN
·Economy·12 min read·WONDY

Backblaze Stock Doubled on the CoreWeave Deal: What the 10-K Says

Backblaze (BLZE) signed a $335 million, five-year storage deal with CoreWeave on June 23 and the stock more than doubled, from around $7.50 in May to about $17.80 by mid-July. I pulled the 10-K and the 8-K from SEC EDGAR and did the math on whether the re-rating holds: the deal versus current revenue, the new valuation, and the scenarios where this breaks.

When a stock doubles in three weeks, what actually changed: the business, or the story about the business? That is the only question worth asking about Backblaze right now. Three weeks ago it was a $449 million cloud storage company most investors had never opened a filing for. Then, on June 23, it announced the largest contract in its history, with one of the hottest names in AI infrastructure, and the market repriced it almost overnight. This is exactly the moment I like to sit down with the actual documents, because a chart moving that fast attracts stories, and stories are where people lose money. So I pulled the 10-K and the new 8-K from SEC EDGAR and did the arithmetic. Here is what the deal actually says, what the re-rating assumes, and where it could all go wrong.

What happened: the CoreWeave deal, in plain terms

First, the company. Backblaze runs two businesses: Computer Backup, a mature flat-fee service that backs up personal computers (grew 3% last year), and B2 Cloud Storage, which rents object storage (the kind of storage apps and AI pipelines use) to businesses at a fraction of big-cloud prices. B2 grew 26% in fiscal 2025 and is the whole growth story.

Now the deal. CoreWeave, the AI cloud provider that rents GPU computing to AI labs, signed a five-year, multi-exabyte agreement under which Backblaze powers the HDD-based storage tiers inside CoreWeave AI Object Storage. Translation: AI training runs need oceans of data parked on cheap disks next to the expensive GPUs, and CoreWeave chose Backblaze to be that parking lot. The master agreement became effective June 16, the announcement hit June 23, and the initial order forms carry an estimated total contract value of about $335 million across five- and seven-year terms, per the 8-K.

The market's reaction was immediate: up about 44% on announcement day to 11.66,thenacontinuedclimbasbrokeragespiledin.Needhamraiseditstargetfrom11.66, then a continued climb as brokerages piled in. Needham raised its target from 8.50 to 14,B.Rileyfrom14, B. Riley from 7 to 16,CraigHallumupgradedfromHoldwitha16, Craig Hallum upgraded from Hold with a 16 target, William Blair upgraded from Underperform. By the second week of July the stock traded around 17.80,nearits52−weekhighof17.80, near its 52-week high of 18.25, putting the market cap at roughly $1 billion.

Backblaze CoreWeave deal infographic — 335 million dollar five-year contract, BLZE stock from about 7.50 to 17.80 dollars, market cap roughly 1 billion
Backblaze after the CoreWeave deal

The math: how big is $335 million for this company?

This is the part a headline cannot tell you and a 10-K can. Backblaze's entire fiscal 2025 revenue was 145.8million,up14145.8 million, up 14% from the prior year. Spread the 335 million contract evenly over five years and you get roughly $67 million a year, equivalent to about 46% of current annual revenue, from a single customer.

That flat-average math is deliberately naive, and I want to be honest about why. The filings do not disclose the revenue ramp: contracts like this typically start smaller and build as capacity deploys, some order forms run seven years not five, and none of it may land evenly. So treat $67 million a year as a rough order of magnitude, not a forecast. Even with that caveat, the conclusion survives: this single agreement is in the same weight class as the company's entire existing business. That is why the stock repriced, and for once the size of the move has a size of a reason behind it.

Hold those two bars in your head: 145.8millionfortheentirecompanylastyear,roughly145.8 million for the entire company last year, roughly 67 million a year from this one agreement. The deal also lands on a financial trajectory that was already improving. Net loss narrowed from 48.5millioninfiscal2024to48.5 million in fiscal 2024 to 25.6 million in fiscal 2025, gross margin reached 61%, and Q1 2026 (reported May 4) delivered 10millionofadjustedEBITDAata2610 million of adjusted EBITDA at a 26% margin with revenue of 38.67 million. One translation note: "adjusted EBITDA" excludes real costs, notably stock-based compensation, and GAAP net income is still negative. Improving trajectory, yes. Profitable company, not yet.

Does the re-rating hold? The honest frame

Before June 23, Backblaze traded around 3 times trailing sales, arguably cheap for a growing cloud business. That discount is gone. At roughly 1billionagainst1 billion against 145.8 million of trailing revenue, the multiple is now about 7 times sales. The market has already paid for a large chunk of the CoreWeave future. So the question is no longer "is this undervalued." It is "did the business change enough to deserve the new price," and that splits cleanly into two readings of the same filings.

The bull reading goes like this. An anchor customer of CoreWeave's profile validates B2 as real AI infrastructure rather than a discount backup vendor; the contract adds revenue comparable to half the existing business, locked in for five-plus years; and it arrived at a company whose losses were already halving on their own. On that reading, 7x sales for a business about to change weight class is not obviously expensive, and the parade of upgrades (Needham, B. Riley, Craig Hallum, William Blair) says professional analysts see the same shape.

The skeptical reading uses identical numbers and lands somewhere colder. At 7x sales, the market is paying today for a ramp whose schedule nobody outside the two companies has seen. The analyst targets that fueled the climb (11to11 to 16) mostly sit below where the stock already trades, which means the crowd ran past the professionals. And a single customer now dominates the growth story of a company that, until three weeks ago, was valued like a slow commodity vendor. Both readings are defensible. That should tell you how much room judgment, not data, occupies here, and why I am not going to pick one for you.

Where this breaks

Three specific failure modes, and I would take them seriously precisely because the chart looks euphoric.

Execution and ramp risk. A multi-exabyte build-out means Backblaze must deploy storage capacity fast and to spec. If deployment slips, revenue recognition slips with it, and a stock priced for smooth execution reacts badly to lumpy reality. The estimated contract value is also just that, an estimate tied to order forms, not revenue in the bank.

CoreWeave concentration. One customer worth roughly 67millionayearagainsta67 million a year against a 145.8 million base is a dependency, not just a win. CoreWeave itself is a heavily investing, debt-financed AI infrastructure company whose own fortunes ride on sustained AI demand. If AI capex cools or CoreWeave rearchitects its storage strategy, Backblaze's growth story loses its biggest chapter. Diversification of the customer base is now a thing to actively verify each quarter, not assume.

The old risks did not leave. Amazon, Google, and Microsoft can still cut storage prices any quarter they like, squeezing that 61% gross margin. And GAAP profitability still has not arrived; if adjusted numbers stay the only good numbers while stock compensation keeps diluting, the re-rating rests on a promise, not a P&L.

What to watch next

  • Q2 2026 earnings (Backblaze has reported in early-to-mid August in past years; the date posts on its IR page). First management commentary on the CoreWeave ramp, and whether B2 growth and the 26% adjusted EBITDA margin held.
  • Revenue recognition disclosures in the next 10-Q on EDGAR: how much CoreWeave revenue lands in 2026 versus later years. This single line will say more than any price target.
  • Additional order forms or new large customers: the difference between "AI storage vendor" and "CoreWeave subcontractor" is whether a second anchor shows up.

How I actually read this: EDGAR plus AI

Everything above came from free public documents, and the routine repeats for any U.S. stock that just made news. Go to SEC EDGAR, search "BLZE," open the fiscal 2025 10-K and the late-June 8-K with the CoreWeave press release. Copy the income statement section and the deal disclosure, then paste both into a chatbot with this prompt:

Below are (A) the income statement section of a company's 10-K and (B) a
press release about a new contract. Explain to a beginner investor.

1. Summarize the revenue and net loss trend in (A), one line each.
2. Compare the contract value in (B) to current annual revenue. Show the math.
3. What does this filing NOT tell us that we would need to judge the deal's
   real impact?
4. Give 3 risks a beginner would miss reading only the press release.

Important: use ONLY numbers present in the text below. If something is not
in the text, say "not in filing" instead of estimating.

[paste 10-K income statement + 8-K press release here]

Here is the core of what came back when I ran Backblaze's documents through it:

AI 답변 (요약)ChatGPT

Trend

  • Revenue: $145.8M in fiscal 2025, up about 14% year over year.
  • Net loss: improved from (48.5)Mto(48.5)M to (25.6)M, roughly half the prior year.

Deal vs. revenue: 335Moverfiveyearsaveragesabout335M over five years averages about 67M per year, equal to roughly 46% of fiscal 2025 revenue. Note this assumes an even split, which the documents do not confirm.

What the filings do not tell us: the revenue ramp schedule, the margin on this specific contract, and any termination or minimum-commitment terms. Without those, the deal's yearly impact cannot be pinned down.

Three risks a press release hides: estimated contract value is not guaranteed revenue; heavy reliance on a single customer's health and strategy; the market may have already priced in the good news, so even solid execution might not move the stock further.

That third answer is why I bother with this routine at all: the AI, boxed to the documents, correctly listed what is missing, and what is missing is exactly where the risk lives. I verified every figure above against the filings and coverage by hand, and if you are considering acting on any of this, do the same. My job ends at showing where the data lives. Whether a doubled stock deserves its new price is a judgment I am deliberately leaving on your desk.

New to reading rate and market data? Start with how the Fed's rate reaches your wallet, then come back to company filings.

FAQ

What is the Backblaze CoreWeave deal?

Announced June 23, 2026: a five-year, multi-exabyte data storage agreement under which Backblaze supplies the HDD-based storage tiers inside CoreWeave AI Object Storage. The master agreement became effective June 16, and the initial order forms carry an estimated total contract value of about $335 million across five- and seven-year terms. It is the largest contract in Backblaze history, disclosed in an 8-K filing on SEC EDGAR.

Why did Backblaze stock double in 2026?

The CoreWeave contract. The stock jumped about 44% on announcement day to 11.66andkeptclimbingasatleastsixbrokeragesraisedpricetargetsorupgradedratings,withnewtargetsbetween11.66 and kept climbing as at least six brokerages raised price targets or upgraded ratings, with new targets between 11 and 16.BythesecondweekofJulyittradedaround16. By the second week of July it traded around 17.80, near its 52-week high of 18.25,foramarketcapofroughly18.25, for a market cap of roughly 1 billion, up from about $449 million in early May.

Is Backblaze still cheap after the CoreWeave deal?

Much less so. Against fiscal 2025 revenue of 145.8million,aroughly145.8 million, a roughly 1 billion market cap is about 7 times trailing sales, versus about 3 times before the deal. The 335millioncontractaveragesouttoroughly335 million contract averages out to roughly 67 million a year, which would be a large addition to current revenue, but the ramp timing is not public. Whether 7x is justified now depends on execution, and that is a judgment this post deliberately does not make for you.

Where can I read the actual contract disclosure?

SEC EDGAR hosts it for free. Search the ticker BLZE, and look for the Form 8-K filed in late June 2026 with the CoreWeave press release attached, plus the fiscal 2025 Form 10-K for the baseline financials. Paste the relevant sections into an AI chatbot for a plain-English readout, then verify every number against the filing text.

Sources