Cerebras Systems files to go public; preliminary S-1 discloses 750MW OpenAI commitment, $1B working capital loan, and 20-to-1 dual-class structure

We estimate that we will receive net proceeds from this

offering of approximately $ (or $ if

the underwriters exercise their over-allotment option in

full), based upon the assumed initial public offering

price of $ per share of Class A common stock,

which is the midpoint of the estimated price range set

forth on the cover page of this prospectus, and after

deducting estimated underwriting discounts and

commissions and estimated offering expenses payable

by us.

Once this offering is completed (and without giving

effect to any shares that may be purchased in this

offering or pursuant to our directed share program), the

holders of our outstanding Class B common stock will

hold approximately % of our outstanding shares

and control approximately % of the voting power

of our outstanding shares, and our executive officers,

directors, and stockholders holding more than 5% of our

outstanding capital stock, together with their affiliates,

will beneficially own, in the aggregate, approximately

% of our outstanding shares and control

approximately % of the voting power of our

outstanding shares.

The holders of our outstanding Class B common stock

will have the ability to control the outcome of matters

submitted to our stockholders for approval, including

the election of our directors and the approval of any

change in control transaction.

We will have three classes of common stock: Class A

common stock, Class B common stock, and Class N

common stock. Each share of Class A common stock is

entitled to one vote per share, each share of Class B

common stock is entitled to 20 votes per share and is

convertible at any time into one share of Class A

common stock, and each share of Class N common

stock is non-voting and is convertible into one share of

Class A common stock.

28,361,707 shares of our Class B common stock issuable upon the exercise of outstanding stock options as

of December 31, 2025, with a weighted-average exercise price of $4.97 per share

Net income (loss) ... . | $237,827 | $(481,602)

Our revenue increased from $24.6 million in 2022

to $78.7 million in 2023 and to $290.3 million in 2024, representing a more than tenfold increase over three years.

Our revenue increased to $510.0 million in 2025, representing year-over-year growth of 76%. We earned net income

of $237.8 million in 2025 and incurred net loss of $481.6 million in 2024.

OpenAI, the world’s leading foundation model lab, selected us

to be its fast inference solution. With Cerebras, OpenAI’s Codex-Spark users turn ideas into working software in

seconds. Amazon Web Services (“AWS”), the world’s leading hyperscale cloud, has signed a binding term sheet

with us to become the first hyperscaler to deploy Cerebras in its own data centers, providing massive distribution to

a broad base of enterprise customers.

our top ten customers by year-to- ... date revenue through December 31, 2025 increased their aggregate spend with us by approximately 80% within 12

months of their initial purchase, often including contracts for co-development.

The WSE-3 is 58 times larger than NVIDIA’s B200

chip. The WSE has 19 times more transistors, 250 times more on-chip memory, and 2,625 times more memory

bandwidth than NVIDIA’s B200 package, which contains two individual chips.

The

combined market for AI training infrastructure and our addressable market within AI inference is estimated to be

$251 billion in 2025 and is expected to grow to $672 billion by 2029—a 28% CAGR, according to Bloomberg

Intelligence.

We estimate that we will receive net proceeds from this offering of approximately $ (or $ if the

underwriters’ over-allotment option is exercised in full), based on an assumed initial public offering price of

$ per share, which is the midpoint of the estimated price range set forth on the cover page of this prospectus,

and after deducting estimated underwriting discounts and commissions and estimated offering expenses payable by

us.

We intend to use a portion of the net proceeds from this offering to satisfy tax withholding and remittance

obligations related to the RSU Net Settlement. Based on the assumed initial public offering price of $ per

share of Class A common stock, which is the midpoint of the estimated price range set forth on the cover page of

this prospectus, an estimated shares underlying RSUs vesting in connection with our initial public

offering, and an assumed % tax withholding rate, we would use approximately $ to satisfy our tax

withholding and remittance obligations related to the vesting of such RSUs.

We currently intend to use the net proceeds

from this offering, together with our existing cash, cash equivalents, and investments, for general corporate

purposes, including working capital, operating expenses and capital expenditures. We may also use a portion of the

net proceeds to in-license, acquire, or invest in complementary technologies, assets, businesses, or intellectual

property.

As of December 31, 2025, our historical net tangible book value (deficit) was $ , or $ per share

of our Class A common stock.

Our pro forma net tangible book value as of December 31, 2025 was $ , or $ per share of

Class A common stock.

After giving effect to (i) the pro forma adjustments set forth above, (ii) the sale by us of shares of our

Class A common stock in this offering at an assumed initial public offering price of $ per share, which is the

midpoint of the estimated price range set forth on the cover page of this prospectus

The foregoing tables and calculations (other than the historical net tangible book value calculation) are based

on no shares of our Class A common stock, shares of our Class B common stock, and no shares of our

Class N common stock outstanding as of December 31, 2025

28,361,707 shares of our Class B common stock issuable upon the exercise of outstanding stock options as

of December 31, 2025, with a weighted-average exercise price of $4.97 per share

Group 42 Holding Ltd

(together with its affiliates, “G42”) accounted for 24.0% and 85.0% of our total revenue for the years ended

December 31, 2025 and 2024, respectively, and in the year ended December 31, 2025, Mohamed bin Zayed

University of Artificial Intelligence (“MBZUAI”) accounted for 62.0% of our total revenue. In December 2025, we

entered into the MRA with OpenAI, which represents a substantial portion of our projected revenues over the next

several years.

We earned net income of $237.8 million and incurred net loss of $481.6 million for the years ended

December 31, 2025 and 2024, respectively, but have a history of generating net losses. We incurred non-GAAP net

loss of $75.7 million in 2025 and $21.8 million in 2024, after excluding the impact of stock-based compensation

expense and change in fair value (extinguishment) of forward contract liability from our GAAP net income (loss).

For more information and for a reconciliation of non-GAAP net loss to net income (loss), see the section titled “— ... Non-GAAP Financial Measures.” As of December 31, 2025, we had an accumulated deficit of $905.3 million.

In addition, under the MRA with OpenAI, unless otherwise agreed to, we are required to deliver capacity

tranches across specified numbers of data centers with minimum capacity thresholds upon certain time-based

milestones. If we fail to deliver such capacity on the stated timelines, or if we experience a certain level of failure

with respect to our service levels, OpenAI has the right to terminate a portion or all of the agreement. In addition,

pursuant to the MRA with OpenAI, we received the $1.0 billion Working Capital Loan (as defined in the section

titled “Management’s Discussion and Analysis of Financial Condition and Results of Operations”). The Working

Capital Loan has a maturity date of no later than December 31, 2032, and is scheduled to be repaid in equal

amortized installments over a three-year term, commencing after the delivery of the final tranche of the initial 250

MW of capacity. Interest accrues on the outstanding principal balance at a rate of 6% per annum; provided that

interest may be waived under certain circumstances. If the MRA is terminated for any reason other than OpenAI’s

material uncured breach, or if certain trigger events occur, OpenAI may direct the bank to cease complying with our

instructions regarding the Working Capital Loan funds and may instead control the disposition of funds in the

account, and we may be required to immediately repay the outstanding principal balance of the Working Capital

Loan together with accrued interest.

With respect to hardware, we compete against semiconductor

companies such as NVIDIA Corporation, a dominant market leader, Advanced Micro Devices, Inc., Intel

Corporation, as well as AI accelerators developed by hyperscalers and private companies. We also compete against

full-service cloud service providers such as Amazon.com, Inc. (AWS), Microsoft Corporation (Azure), Alphabet

Inc. (Google Cloud Platform), and Oracle Corporation, as well as AI-optimized specialized clouds such as

CoreWeave, Inc. and other neo-clouds. Certain of our competitors are also current or prospective customers. Many

of these companies are large, well established, and have greater financial resources than we do.

We depend on one third-party foundry, Taiwan Semiconductor Manufacturing Company Limited

(“TSMC”), to manufacture our proprietary processor using its fabrication equipment and techniques.

Substantially all of our manufacturing services and component orders are currently transacted

on a purchase order basis, and in many instances we negotiate pricing separately for each purchase order, with our

contractors and suppliers having no obligation to perform services or supply products to us for any specific period,

in any specific quantities, or at any specific price, except as may be provided in a particular purchase order.

we released our inference solution in 2024,

and if we fail to win broad customer adoption, we will be unable to participate in a significant segment of the AI

market.

Further, in 2025,

a significant majority of our revenue was generated from customers headquartered in the United Arab Emirates.

In December 2025, we entered into a Master Relationship Agreement (the “MRA”) with OpenAI OpCo, LLC

(“OpenAI”), under which OpenAI agreed to purchase 750MW of AI inference compute capacity (the “Committed

Capacity”) and related services, including collaboration on engineering development and integrations. ... In addition, to accelerate our

engineering development, manufacturing scale up and data center expansion, OpenAI advanced to us a working

capital loan of approximately $1.0 billion (the “Working Capital Loan”), which we will repay in cash or through the

delivery of compute capacity or hardware or other services under the MRA. The Working Capital Loan is subject to

a secured promissory note with a maturity date of no later than December 31, 2032, which may be prepaid in whole

or in part at any time without penalty, and which bears interest at 6% per annum, provided any accrued interest shall

be waived to the extent such loan is repaid through delivery of compute capacity, hardware or other services, or

assignment or transfer of certain assets, in accordance with the MRA.

In December 2025, we issued the OpenAI Warrant to OpenAI in connection with the execution of the MRA.

Pursuant to the OpenAI Warrant, OpenAI has the right to purchase up to 33,445,026 shares of our Class N common

stock at an exercise price of $0.00001 per share. The OpenAI Warrant expires on the earlier of December 24, 2035

and five business days following the first date during which there is no binding capacity purchase commitments or

contractually obligated current or future payments under the MRA. The shares of Class N common stock underlying

the OpenAI Warrant vest upon the achievement of certain commercial and market capitalization milestones, with

full vesting occurring only if OpenAI exercises all options to purchase Additional Capacity under the MRA, such

that a total of 2GW of AI inference compute capacity and related services is purchased by OpenAI.

For the year ended December 31, 2025, MBZUAI

accounted for 62% of our total revenue. G42 accounted for 24% and 85% of our total revenue for the years ended

December 31, 2025 and 2024, respectively.

Our revenue increased to $510.0 million in 2025, representing year-over-year growth of 76%. Our gross margin was

12%, 33%, 42% and 39% in 2022, 2023, 2024 and 2025, respectively. We earned net income of $237.8 million in

2025 and incurred net loss of $481.6 million in 2024.

However,

gross margin is expected to initially be lower compared to recent periods, and may fluctuate, due to factors such as

contra-revenue that will be recognized in future periods in connection with warrants granted to customers that will

lower the amount of revenue reported compared to what would have been recognized without such warrants,

customer arrangements with pass through amounts for data center costs that will be included in both revenue and

cost of revenue and will have a dilutive impact on gross margin, and other start-up costs related to expediting the

availability of cloud capacity to fulfill the significant increase in near-term demand. When we begin to recognize

contra-revenue amounts from the warrants in the first quarter of 2026, we expect quarterly revenue growth rates will

decline from recent trends.

Year Ended December 31, | 2025 | 2024 | $ Change | % Change | (in thousands, except percentages) | Hardware ... . | $358,440 | $211,965 | $146,475 | 69% | Cloud and other services ... 151,551 | 78,287 | 73,264 | 94% | Total revenue ... . | $509,991 | $290,252 | $219,739 | 76%

Year Ended December 31, | 2025 | 2024 | $ Change | % Change | (in thousands, except percentages) | Gross profit ......................................................... | $199,071 | $122,738 | $76,333 | 62%

Gross margin .......................................................

39.0%

42.3%

Other income (expense), net for the year ended December 31, 2025 increased by $769.0 million, or 203%,

compared to the year ended December 31, 2024. The increase in other income (expense), net was primarily

attributable to a gain of $363.3 million recognized upon extinguishment of the forward contract liability during

2025, compared to remeasurement losses of $401.3 million on the forward contract liability in 2024.

Our revenue increased from $24.6 million in 2022

to $78.7 million in 2023 and to $290.3 million in 2024, representing a more than tenfold increase over three years.

Our revenue increased to $510.0 million in 2025, representing year-over-year growth of 76%. We earned net income

of $237.8 million in 2025 and incurred net loss of $481.6 million in 2024.

OpenAI, the world’s leading foundation model lab, selected us

to be its fast inference solution. With Cerebras, OpenAI’s Codex-Spark users turn ideas into working software in

seconds. This partnership is an example of tight hardware-software co-design with a leading frontier model lab.

AWS, the world’s leading hyperscale cloud, has signed a binding term sheet with us to become the first hyperscaler

to deploy Cerebras in its own data centers, providing massive distribution to a broad base of enterprise customers.

We signed the MRA with OpenAI on December 24, 2025. On January 23, 2026, we began delivering

capacity to OpenAI, and on February 12, 2026, OpenAI’s Codex-Spark model, powered by Cerebras infrastructure,

was made available to the public. Spark is OpenAI’s model designed for real-time coding. Using Cerebras,

OpenAI’s customers can translate ideas into working software in seconds, enabling developers to create software at

the speed of thought. OpenAI has committed to purchase 750MW of Cerebras inference compute capacity over the

next three years.

We signed a binding term sheet with Amazon Web Services for AWS to become the first hyperscaler to

deploy Cerebras systems in its data centers. Deployment in AWS data centers will require us to meet strict standards

for performance, scale, and reliability.

Pursuant to the term sheet, we will create a co-designed, disaggregated inference-serving solution that will

integrate AWS Trainium3 chips with Cerebras CS-3 systems, connected via high-bandwidth networking, to partition

inference workloads across Trainium3 and CS-3. Each system will perform the type of computation at which it most

excels. The approach is expected to deliver 5 times more token throughput in the same hardware footprint, at up to

15 times faster speeds compared to leading GPU-based solutions as benchmarked on leading open-source models.

Our product portfolio spans on-premises AI supercomputing systems, cloud-based compute for training and inference, and forward-deployed AI services to help customers accelerate the creation and deployment of AI capabilities. On-premises customers procure our AI supercomputers through a traditional purchase-order process with payment received upon delivery or acceptance. Each system combines tightly integrated hardware and software and the purchase includes a separate renewable software subscription for continuous updates and upgrades, generating a recurring revenue stream. Customers procure cloud capacity from us and our cloud partners through two primary models: Dedicated Capacity and On-Demand. Dedicated capacity provides availability and is well suited for production deployments and large-scale workloads. These contracts are generally structured as take-or-pay commitments, under which customers pay for dedicated compute capacity irrespective of utilization.

Our gross margin was 12%, 33%, 42%,

and 39% in 2022, 2023, 2024, and 2025, respectively.

our top ten customers by year-to- ... date revenue through December 31, 2025 increased their aggregate spend with us by approximately 80% within 12

months of their initial purchase, often including contracts for co-development.