Skip to content 
In December 2020, C3.ai was a poster child of the AI revolution. It listed at $42 and surged to $177 within weeks, riding the wave of hope that it could bring artificial intelligence to every enterprise.
Fast forward nearly five years, and the stock is trading around $24 (see Figure 1).
It is not that the company stood still. C3.ai signed billion-dollar defense contracts, launched enterprise-grade generative AI tools, and brought in heavyweight partners like Microsoft, AWS, and PwC.
Yet, the market remains skeptical.
The reason? It still loses more money than it earns. In FY25, it generated $389 million in revenue but with a loss of $289 million at a net profit level. Basically, the company spends more than $1.44 for every $1 it earns. That kind of math tests investor patience.
But here is where it gets interesting: Product launches in generative AI are gaining traction across industries. And for C3.ai, over 70% of new deals now involve major partners like Microsoft, AWS, and PwC, indicating a scalable sales model is starting to take shape.
So, now the question is this: if the business is evolving, why has the stock stayed stuck?
In this article, we break down the numbers, strategy, risks, and potential scenarios for C3.ai and whether the next move is finally worth watching.
Figure 1: Share Price of C3.ai. Source: Vested Finance
What Exactly Does C3.ai Do?
This is not a company selling AI chips or infrastructure. It is not building a consumer-facing chatbot either. C3.ai is focused on enterprise software – specifically, helping large organizations use artificial intelligence to improve operational efficiency, reduce downtime, and make smarter decisions using the data they already have.
In simple terms, C3.ai develops AI tools for real-world problems, serving clients (see Figure 2) such as the U.S. Air Force, Shell, Dow Chemicals, and state governments.
Figure 2: Case Studies Shared by the Company. Source: Official Website of C3.ai
Now here is where it gets interesting. C3.ai is not a general-purpose AI company. It is a platform plus apps business.
Think of it as a three-layer model.
- The Platform (Agentic AI): At the foundation is the C3 Agentic AI Platform, a full-stack environment where developers can build AI applications.
In theory, enterprises can license just the platform and use it to create custom AI agents that process data, run predictions, and even take decisions autonomously. This seems powerful, but in FY25, only 5% of C3.ai’s bookings came from customers buying just the platform.
Why? Because most companies do not want to build from scratch. They want ready-made tools.
- The Applications: That brings us to the core of C3’s business – prebuilt AI applications.
These are software products designed for specific use cases: predictive maintenance, fraud detection, energy optimization, inventory planning, and more.
For example, the U.S. Air Force uses C3.ai’s predictive maintenance software (under the PANDA project) to anticipate aircraft part failures.
In FY25, 95% of C3.ai’s bookings came from these ready-to-use AI applications, which shows that enterprises prefer solutions over platforms. The applications run on top of the platform but are faster to deploy and offer direct business value.
- The Generative AI Layer: In FY25, C3.ai added a third layer, C3 Generative AI.
This is where things start to feel modern. These tools allow employees in large organizations to ask questions of their internal data in natural language, just like they would chat with ChatGPT, but behind a secure firewall, and using enterprise data.
So, a maintenance planner in a defense unit can ask, “Which aircraft part is most likely to fail in the next 30 days?” and get an explainable answer backed by historical data.
In FY25, C3.ai signed 66 deployment agreements for Generative AI across 16 industries. Revenue from this category doubled over the previous year. And because these tools are embedded across every app C3.ai offers, they are designed to scale quickly.
The Business Model: Mostly Software, Mostly Recurring
One of the clearer takeaways from FY25 is that C3.ai is, in structure, a software-as-a-service (SaaS) company:
- 84% of its FY25 revenue (vs. 89% in FY24, see Figure 3) came from subscriptions, meaning customers pay over time to use the software.
- The remaining 16% came from services like onboarding, training, and what C3 calls “prioritized engineering” essentially paid customizations that often become part of the main product later.
Figure 3: Company’s Revenue Split. Source: Year End Press Release
But there is a shift happening. Earlier, deals were large and long-term as customers would sign multi-year contracts for big-ticket deployments. Now, C3.ai seems to be leaning into shorter, smaller initial deployments, what it calls “initial production agreements.”
In FY25, it signed 174 of these, up 41% from the previous year (see Figure 4). The average deal size is smaller, but the volume is higher.
Figure 4: Management on Agreement Closures. Source: Year End Press Release
Why is this important?
Because it reflects a “land and expand” strategy: get inside more organizations with a narrower use case, prove value, and then grow from there. It lowers the barrier to entry for customers and gives C3.ai more at-bats. But it also delays revenue recognition and makes near-term growth less predictable.
But Then, Who Is Buying This Software?
C3.ai is not selling to startups or tech-native firms. Its focus is on large, traditional organizations with complex operations and underused data.
In FY25, its customers spanned 19 industries, with clear concentration:
- Federal and Defense (26%): Major contracts with the U.S. Air Force, Navy, and intelligence agencies, including a $450 million PANDA deal through 2029.
- Oil and Gas (19%): Continued sales via Baker Hughes to Shell, ExxonMobil, and others, though this vertical is declining in share.
- Manufacturing and Chemicals (12%): Clients like U.S. Steel and Dow use C3.ai to improve reliability and efficiency.
- State and Local Governments (7%): Adoption doubled, with over a dozen states using AI for tax, fraud, and planning.
- Others: Pharma, agriculture, and high-tech are emerging areas, with new wins like Bristol Myers Squibb and Chanel.
This is not easy-to-sell consumer software. It is enterprise-grade, often mission-critical, and designed for long-term use.
Figure 5: Revenue Spread by Sectors. Source: Company’s Q4FY25 Report
Margins and Valuation: Can the Business Justify the Price?
One of the most important questions investors ask about a company like C3.ai is simple: How much is it burning, and how much is it earning?
Because no matter how promising the technology, if a business cannot convert revenue into profit, the market eventually stops rewarding growth alone.
What the Margins Tell Us
C3.ai closed FY25 with $389 million in revenue, up 25% year-on-year. On the surface, that looks healthy. But dig a little deeper, and two things stand out:
- Gross margin came in at 69.6% – a strong number, and typical for a software company. This means C3.ai’s core product, its platform and AI applications, scales well. Once the cost of running the software is accounted for, the company retains nearly 70 cents of every dollar earned. That is a good starting point.
- Operating losses, however, remain significant. The company posted an operating loss of $324 million and a net loss of $288 million. Free cash flow for the year stood at -$44 million.
So what is happening?
C3.ai is spending heavily on R&D and go-to-market. It is trying to stay ahead in a fast-evolving field, while also acquiring new customers and expanding existing accounts. The company has made it clear: profitability is not the goal today, but growth and product leadership are.
This is not unusual for a company in this stage. But it does mean that investors have to be patient, and they have to believe that at some point, these losses will narrow as scale kicks in.
So, What Are Investors Paying For?
As of June 2025, C3.ai trades at around $24 per share, with a market cap of $3.2 billion.
On FY25 revenue of $389 million, that puts the price-to-sales (P/S) ratio at 8.2x, a premium for a company that is still unprofitable, early in monetizing generative AI, and facing larger rivals.
So, why the premium valuation? The market appears to be pricing in cautious belief, not hype. There is optimism that with growing federal contracts and strategic partnerships, the business could scale into its valuation.
C3.ai expects FY26 revenue of $447–485 million. At the midpoint ($466 million), that implies 20% growth and a forward P/S of ~6.9x. More reasonable but only if growth holds and losses narrow. If not, multiple compression is likely.
A Simple Way to Think About the Valuation
Let us break it down into three possible paths from here:
1. The Upside Case
If C3.ai grows at ~25% CAGR for the next two years and manages to cut losses meaningfully, the market could reward it with a higher multiple, say 10x forward sales, especially if generative AI adoption accelerates across industries.
- FY27 Revenue: ~$610 million
- At 10x: Market cap could reach $6 to 6.5 billion
- That implies a stock price of around $45 to 50
However, this depends on several things going right, including strong federal execution, meaningful enterprise expansions, and operating leverage starting to show.
2. The Base Case
Growth continues at ~20%, margins improve slowly, but competition remains stiff. The market maintains a P/S multiple around 6–7x.
- FY27 Revenue: ~$565 million
- At 6.5x: Market cap remains around $3.6 to 3.8 billion
- That means a stock price in the range of $28 to 30
This would imply moderate upside, provided execution remains consistent and investor confidence is sustained.
3. The Downside Case
Growth slows to mid-teens or lower, and losses persist longer than expected. Investors begin to demand profitability before assigning high valuations.
- FY27 Revenue: ~$500 million
- At 4- 5x: Market cap could fall to $2 to 2.5 billion
- That would put the stock closer to $15 to 18
This scenario could play out if major clients scale back, if federal budgets tighten, or if competitors (like Microsoft or Palantir) win away key contracts.
The stock has stayed flat for over a year. That signals the market is waiting for evidence: of margin expansion, deal scale-up, and long-term customer retention. Until then, C3.ai may remain stuck between potential and proof.
