1. Snapshot – Where Palantir Stands Going into 2026

Palantir Technologies sits in a pretty unique spot in early 2026: an AI software company that still gets a very large chunk of its money from government and defense, but whose fastest-growing engine is now U.S. commercial AI. The latest full reported quarter (Q3 2025) showed total revenue of roughly $1.18 billion, up about 63% year over year and strong profitability, with management raising full-year 2025 revenue guidance to around $4.396–4.400 billion. The real shock, however, is the mix:

U.S. commercial revenue grew roughly +121% year over year in that quarter, reaching around $397 million, while U.S. government revenue still grew a solid +52% to about $486 million. The CEO Alex Karp has been calling the commercial engine an “absolute juggernaut” and describing demand for the AI Platform as “otherworldly”, and the numbers back up the claim.

2. Business Model – From Gotham to AIP

Palantir’s original DNA is government analytics: Gotham and related platforms built for intelligence, counterterrorism, law enforcement and military users. Over the last years, the company has layered on Foundry (data integration and analytics for enterprises) and then the AI Platform (AIP), which lets customers orchestrate and deploy large language models and other AI models directly on top of their own data.

In practical terms, Palantir sells more than “just software licenses”. The model combines:

Long-term contracts with governments, often multi-year and milestone-based, for platforms that become deeply embedded in operations (logistics, targeting, battlefield awareness, public health, immigration enforcement and more).

Commercial deployments where Palantir runs “bootcamps” and high-touch pilots to show value with AIP in weeks, then tries to expand usage across factories, hospitals, banks, insurers and critical infrastructure.

The flip in 2024–2025 is that the commercial engine finally started to outrun government in terms of growth rate, with U.S. commercial becoming the fastest-growing piece of the entire company.

3. Defense Backbone – The U.S. Army Enterprise Agreement

On 31 July 2025 the U.S. Army announced a new Enterprise Agreement (EA) with Palantir. The idea is simple but huge: consolidate roughly 75 existing contracts into one 10-year umbrella agreement with a maximum potential value of up to $10 billion over a decade. It is important to stress that this is a ceiling, not guaranteed revenue; however, it creates a streamlined path for the Army and parts of the Department of Defense to buy Palantir software and services with volume-based discounts.

Army and defense officials frame the EA as a way to reduce friction, cut intermediary fees and standardize on a few core platforms for data, AI and analytics. For Palantir this means:

A decade-long, visible landing zone inside Army modernization, complementary to existing programs like Army Vantage and other battlefield analytics platforms.

A strong signal to other U.S. and allied agencies that Palantir is “safe enough” and battle-tested enough to be treated as a strategic vendor for AI and software.

From an equity angle, the EA does not automatically create $1 billion per year of revenue, but it anchors the defense leg of the story and makes the government pillar stickier and more predictable over time, as long as budget and political conditions cooperate.

4. Commercial AI – AIP as the Growth Engine

The surprise of 2025 has been how fast the commercial AI engine has scaled. In Q3 2025, U.S. commercial revenue grew roughly +121% year over year, to about $397 million in a single quarter, compared with about +52% for U.S. government. Management repeatedly highlighted:

A record “Rule of 40” score of more than 100, combining fast revenue growth with high operating margins.

A pipeline of AI Platform (AIP) deals across sectors as different as manufacturing, healthcare, energy, financial services and logistics.

The strategy is to sell AIP not as a generic chatbot, but as a layer that sits directly on top of real operational data, allowing companies to simulate scenarios, ask questions, trigger workflows and even control robots or industrial systems through AI agents. The Innodata partnership connects directly into this, because it is about making sure the underlying multimodal data (video, images, sensors) is structured and labeled well enough to let AIP do something meaningful in the real world.

5. Innodata Deal – Why Rodeo Analytics Actually Matters

At the end of January 2026, Innodata announced that it had been selected by Palantir to support “advanced initiatives in AI-powered rodeo modernization”. Behind the colourful language there is a concrete technical job: Innodata will provide specialized annotation and data engineering for thousands of hours of rodeo video footage, enabling computer vision models to:

Detect animals and riders, track skeleton joints and calculate performance metrics in bull riding, bronc riding, bareback riding and barrel racing.

Integrate video with other event and sensor data so that Palantir’s platforms can offer real-time analytics for rodeo organisers, broadcasters, sponsors and regulators.

The partnership sits inside Palantir’s broader expansion of event analytics and multimodal AI use cases: not just text and tables, but video, imagery and sensor streams. From Innodata’s point of view, this is a high-profile validation of their ability to handle complex training data, and the stock has reacted accordingly, with a strong rally since the deal was disclosed.

Financial terms have not been disclosed, so this is not a confirmed “mega-contract” for Palantir in dollar terms. However, strategically it signals a few important things:

Palantir is willing to rely on specialized partners for the “heavy lifting” of labeling and preparing complex multimodal data, instead of trying to do everything in-house.

The same toolkit that is used on rodeo video can, in principle, be applied to far more sensitive domains: drone footage, satellite imagery, border security, smart infrastructure and other defense- or space-related applications.

For the AI investor, the Innodata deal is less about near-term revenue and more about ecosystem: Palantir is building a network of domain and data partners around AIP, which can make the platform stickier and more useful over time.

6. Financial Profile – Growth, Profitability and Concentration

Palantir’s latest reported quarter showed a business that is no longer “growth at any cost”. With revenue up roughly 63% year over year to about $1.18 billion in Q3 2025, the company also delivered strong net income and operating margins, prompting yet another guidance raise for full-year 2025.

Some key points from recent numbers and guidance:

Full-year revenue expected around $4.396–4.4 billion, with U.S. commercial revenue expected to more than double year over year.

Net income margin already solidly positive, helped by scale and by the high gross margins typical of software.

Contract value and remaining deal value growing faster than revenue, which signals a strong pipeline of signed deals not yet fully recognized as revenue.

On the risk side, there are some pressure points:

Palantir invests heavily in marketable securities and growth initiatives, which can move the cash and liquidity line quarter by quarter.

Accounts receivable have been trending up, reflecting rapid contract growth but also adding a layer of working-capital risk that investors need to watch.

7. Risk Map – Valuation, Politics and Contract Risk

Any honest deep dive on Palantir has to deal with the risk side. Beyond standard execution risk, three themes stand out:

Valuation and sentiment

After a huge run in 2025, Palantir’s stock has been under pressure into early 2026. Investors are trying to reconcile very strong growth and margins with a valuation that already prices in years of success. The stock is popular with retail investors and momentum traders, which can amplify both upside and downside moves around earnings and news.

Concentration in government and defense

Even with the commercial AI engine accelerating, Palantir still depends heavily on government and defense contracts across the U.S. and allies. The new U.S. Army Enterprise Agreement underscores that dependence: a strategic positive in terms of staying power, but also a risk if budget priorities or political winds shift.

Historically, Palantir has benefited from waves of spending linked to counterterrorism, border control and now great-power competition. Those flows can be powerful tailwinds, but they are not guaranteed and can become politically controversial.

Reputational and policy risk

Palantir’s work with immigration enforcement, surveillance and military targeting systems has already attracted criticism in parts of the tech community and civil-liberties groups. A shift in public perception or regulation around AI, privacy and military AI could change how easy it is for Palantir to win some types of contracts, especially in Europe.

8. 2026 Catalysts and Timeline – What to Watch

Looking ahead, the Palantir story in 2026 will revolve around a few tangible checkpoints rather than a single “binary” catalyst:

Q4 2025 earnings and first 2026 guidance

The next earnings release is expected to show whether the 60%+ growth profile can be sustained, and how management frames 2026 revenue and margin guidance. The market is watching especially for:

The growth rate of U.S. commercial AI revenue and the number of AIP-driven deals.

Early revenue contribution visible under the U.S. Army Enterprise Agreement and other federal AI programmes.

Army Enterprise Agreement ramp

Throughout 2026 investors will try to track how much actual revenue flows through the 10-year Army EA. Because it is a ceiling, not a guaranteed amount, the ramp path matters more than the headline number. Each new task order or expansion inside the Army and across DoD components helps de-risk the long-term government pillar.

Commercial AI adoption

On the commercial side, the key question is whether Palantir can keep doubling U.S. commercial revenue for another year, or if growth moderates. Concrete datapoints to watch include:

New logos and reference customers in heavily regulated industries (healthcare, finance, energy).

The size and duration of new AIP deals, especially if multi-year commitments start appearing more often.

Innodata and broader ecosystem

The Innodata partnership is one visible piece of a broader ecosystem strategy. Investors will be watching:

Whether Palantir announces additional data/annotation partnerships in areas like satellite imagery, robotics or defense imagery.

Whether Innodata itself discloses follow-on work with Palantir beyond rodeo analytics and any work linked to federal or missile defense programmes.

9. Sentiment – Between AI Darling and Overcrowded Trade

Palantir has become one of the most polarising AI names in the market. On one side, there is a strong base of retail investors and some institutional bulls who see a rare combination of:

High growth, profitability, exposure to both commercial AI and defense, and a very long runway of federal digital-transformation work.

On the other, critics argue that after such a big share-price run, Palantir is priced for near-perfection and carries non-trivial political and contract-concentration risk.

For traders and long-term investors alike, that means volatility around earnings, big contract headlines (like the Army EA) and stories like the Innodata deal is likely to stay high. The core of the investment case going into 2026 is simple to state but hard to execute: can Palantir sustain high-double-digit growth in commercial AI while deepening and broadening its government footprint without hitting a wall in valuation or politics.

10. Key Sources and Further Reading