The Rubber Meets the Road

Market Overview

Markets were slightly weaker this week with both the S&P 500/Nasdaq 100 down roughly 1%/2%. Technology names slightly underperformed on jitters around the cannibalization of existing revenue streams by AI YTD. The CBOE Volatility Index (the “fear index”) has increased to ~21 from ~15, a level that is not yet alarming but is at the high end of “not worried.” Our NPM Private Market Tracker*, which shows the average price performance of the 50 largest names in our internal Tape D® data, is up 2.5% YTD. (Bloomberg; NPM)

LLMs: Strategy, Scale and Staying Power

In our view, 2026 roughly marks the transition of the LLM market into a new phase of competition defined less by growth rate and more by strategy and monetization. While the main story of the “first inning” of AI (i.e. 2023-25) was viral consumer adoption and OpenAI’s first mover advantage, the competitive field in 2026 is now more balanced. For example, Google has now asserted itself as a top tier frontier model provider, and Anthropic has established meaningful enterprise leadership. And in a recent strategic “zag”, xAI was acquired by SpaceX, presumably giving the company access to greater financial resources and even outer space. Open source challengers such as Meta’s Llama and Mistral have also gained structural importance. Finally, emerging “efficiency” players such as DeepSeek have demonstrated that high performance can be achieved with capital light approaches. (NPM)

At this point, from a raw performance perspective, the top frontier models—OpenAI’s GPT, Google’s Gemini, Anthropic’s Claude and xAI’s Grok — are broadly comparable across many standardized benchmarks. Thus, strategy and positioning, in our view, are becoming more important differentiators. (NPM; Bloomberg)

OpenAI remains the category defining brand and still leads in broad consumer usage. However, the company faces increasing strategic pressure. As mentioned, performance differentials among frontier models have narrowed. Second, pricing competition has intensified as open source and efficiency focused models drive down inference costs. Third, OpenAI’s capital intensity remains high, requiring continuous access to financing, and making the MSFT relationship almost mission critical. While the company retains strong brand equity, its long-term trajectory will depend on converting usage dominance into durable revenue streams with defensible margins. To this end, OpenAI is just beginning to roll out advertisements to free users in the US. (NPM)

Google’s AI ascendancy is one of the most important developments in the past 12 months. Although initial adoption lagged OpenAI’s, Google’s Gemini family has emerged as highly competitive. In addition, Google has a structural advantage given that Gemini is embedded across the Google ecosystem, including almost 4bn Android phones. Unlike the standalone LLM vendors, Google is relying on Gemini to enhance existing customer experience and revenue generation vs. utilizing models as discrete revenue streams. (NPM)

Anthropic has also emerged as a formidable competitor in enterprise. Claude’s recent iterations have been particularly strong in structured reasoning, coding and applications for verticals such as legal and financial services. Thus, to some degree, we see Anthropic as the mirror image of OpenAI, which currently is more focused on consumers. While it is still early, we continue to view enterprise AI spend as potentially the larger prize over time, likely commanding higher multiples given a more recurring nature. (NPM)

SpaceX’s Feb 26 acquisition of xAI similarly reflects both competitive and strategic pressures. As the market broadly becomes more concerned about the limits of terrestrial data centers, Musk has argued that moving AI compute into space via Starlink satellites and orbiting data centers could offer vastly more efficient architectures that terrestrial rivals will not be able to readily replicate. The acquisition also helps address the massive capital demands of training and running advanced models in the face of competitors with deep pockets (OpenAI backed by Microsoft, Google with Gemini, etc.), by giving xAI access to SpaceX’s cash flow and proceeds from a possible IPO. (NPM; Bloomberg)

Looking ahead through 2026, we think several themes will emerge. First, differentiation will shift further from “parameter counts” to features like workflow integration and agentic capability. Second, cost-performance economics will become more of a battleground. At some point, as inference costs decline and efficiency improves, pricing pressure is likely to intensify. Third, enterprise procurement patterns will mature, favoring vendors with differentiated offerings. Finally, regulatory developments—particularly around safety standards, and export controls—may shape regional competitive dynamics. (NPM; Bloomberg)

The final, and perhaps most relevant to this audience, topic to mention is valuation. While valuations for the major LLMs largely moved in lock step in 2025, we see potential for greater valuation dispersion to emerge. Anthropic raised capital in early 2025 at a $62bn EV and is currently in discussions to raise $35bn at $380bn (6.1x) post-money. OpenAI was at $157bn and is now reportedly raising capital at up to $830bn (5.3x). xAI entered 2025 at a $50bn valuation and was acquired by SpaceX for $250bn (5x) in Feb ‘26. The law of large numbers dictates that it will be nearly impossible for valuations to scale at the same rate in ‘26 as in ‘25. We also see potential for the parallel moves in LLM valuations to decouple over time as strategic differentiation becomes more prevalent. (NPM; Pitchbook)

RECENT EVENTS

• Anthropic announced it will absorb electricity price increases caused by its data centers in affected areas. (StrictlyVC; 2/11)
• Activist investors are urging Warner Bros. Discovery to abandon its $72bn deal with Netflix in favor of a $78bn offer from Paramount Skydance. (StrictlyVC; 2/11)
• OpenAI started testing ads in the US on ChatGPT (StrictlyVC; 2/9)
• Stripe is considering a tender offer that would value the company at $140bn. (StrictlyVC; 2/12)

NOTABLE CAPITAL RAISES

• Alphabet raised $32bn in debt to help fund AI capex. (Bloomberg; 2/9)
• Humanoid robot maker Apptronik raised $925mm Series A from Google, Mercedes-Benz and B Capital. (Bloomberg 2/11)
• Runway, a NY startup building AI video generation tools, raised a $315mm Series E at a $5.3bn post money valuation. (Bloomberg; 2/10)
• Anthropic raised $30bn at a $380bn post money value. (Bloomberg; 2/12)
• Vista Equity Partners and Intel invested $350mm in AI chip maker SambaNova Systems following a stalled sale of the company. (Reuters; 2/6)
• AI chip maker Cerebras Systems raised $1bn in a round led by Tiger Global that values the firm at $23bn. (Reuters, 2/6)

NOTABLE EXITS

• Broker Clear Street Group postponed its IPO due to market conditions. (Bloomberg; 2/12)