International Business Machines (NYSE:IBM) has had a solid start to 2025. Shares are up 15% as of the Feb. 6 close. This was mostly due to its recent earnings report, which surpassed expectations. IBM, a lesser-talked-about player in AI, is positioning itself as a formidable force.

After IBM’s Q4 2024 earnings results, shares of the tech company jumped by around 13% in one day. On revenue, the results came in essentially in line with estimates at $17.6 billion, exhibiting growth of only 2%. The company’s software segment led the way, with sales rising by 11%. Its software solutions focus largely on helping manage applications that exist in hybrid cloud environments. On earnings per share (EPS), the company solidly beat estimates by nearly 4% as margins in its software segment increased by 220 basis points.

In consulting, the firm works with businesses to figure out how they can tangibly implement GenAI to drive improved performance. Overall, the company’s consulting business grew by only 1% in the quarter. The company attributed this to the fact that customers are “reprioritizing their IT spending towards digital transformation and AI initiatives." Customer spending is shifting toward AI. This change is likely hurting IBM's non-AI consulting business. Those changes in spending have yet to result in a strong increase in overall consulting revenue. This segment should return to growth as the company is able to recognize its billions in bookings as actual revenue.

The rapid growth in the company’s GenAI bookings creates excitement when thinking about the possible flywheel effect from that. To me, consulting on how to use GenAI should be an area where businesses continue to invest heavily over time. The understanding of what AI can do, especially from a business perspective, is still evolving. IBM notes that 98% of businesses have experimented with AI, but only 26% have scaled these experiments into production. As the AI revolution matures from its early stages, IBM has a large opportunity to serve clients' needs.

The company also believes that the emergence of DeepSeek validates its approach to language models. The company views the smaller distilled DeepSeek models as similar to the strategy it has been using to develop its own models. The company creates “built-for-purpose” models. They excel in specific use cases instead of aiming for general capabilities. It compared the performance of its Granite seven billion parameter model to that of Meta Platforms’ (NASDAQ:META) Llama 3 70 billion parameter model. IBM claims that when tasked with a specific business use case, accounting, its model performed 6% better while costing 75% less. This is compelling. To me, this targeted strategy can meet customer needs at a more competitive price.

At this point, IBM is trading at its highest forward price-to-earnings (P/E) multiple over the past 20 years of nearly 24x. That’s slightly below the overall tech sector, but the company is currently only growing at 2%. The company’s highest price target from analysts at Wedbush shows 18% upside in the stock. To me, the story for IBM in AI is there. The possible flywheel's further penetration into the space could be exciting. I would like to see the valuation come down, but IBM stock could continue winning even at these levels.