What DeepSeek’s Rise Reveals About Enterprise AI Adoption
DeepSeek’s rise as a DeepSeek AI alternative highlights how enterprises are reconsidering cost-effective AI models as they embed generative tools into daily operations, forcing teams to weigh cheaper AI solutions against data residency, security controls, and the risk of adding a new strategic vendor to their stack. This shift is visible in Ramp’s June trending software vendors list, where DeepSeek ranks first based on new corporate spending. Ramp’s index tracks first-time vendor purchases, so the ranking signals momentum in enterprise AI adoption rather than market dominance. OpenAI and Anthropic still lead overall spending-based adoption, but their success has made AI a large recurring expense. As token usage grows across coding, customer support, and analytics, more procurement teams are treating DeepSeek’s lower pricing as a serious option, even if it means sending prompts and outputs through an external hosted service instead of keeping every interaction on in-house infrastructure.
Rising AI Costs Push Firms Toward Cheaper AI Solutions
Many enterprises started with small language model pilots but are now running AI in production for software development, customer operations, and analytics. As usage scales, spending on premium models has become hard to ignore, driving interest in cheaper AI solutions that still meet basic performance needs. Ramp’s data shows that this pressure is powerful enough to shift behavior: DeepSeek topped its June trending list, beating better-known providers in terms of fresh vendor trials. According to Ramp Economics Lab, the platform now reflects expense data from roughly 70,000 businesses, giving its rankings weight as an early signal of changing procurement habits. Meanwhile, earlier adoption figures underline the gap: in April, Anthropic and OpenAI accounted for 34.4% and 32.3% of Ramp’s AI adoption index, while DeepSeek sat near 0.1%, underscoring that enterprises are testing alternatives, not overthrowing incumbents.
Why DeepSeek Looks Like a Cost-Effective AI Model
DeepSeek’s main appeal is its promise of cost-effective AI models that can handle common enterprise tasks without the price tag associated with leading players. The startup offers open-source models that companies can self-host, as well as a hosted service that removes infrastructure overhead. Ramp’s June analysis notes that some firms are paying DeepSeek directly instead of only downloading model weights, a sign that buyers see enough value in the managed service to make it part of their enterprise AI adoption strategy. Proactive has reported that DeepSeek is raising USD 7.4 billion (approx. RM34.0 billion) in its first external funding round at a valuation between USD 52 billion (approx. RM239.2 billion) and USD 59 billion (approx. RM271.3 billion), a war chest that could support infrastructure build-out and product improvement. For many buyers, that financial backing makes DeepSeek look more like a long-term partner than a short-lived budget option.
The Data Residency and Security Trade-Off
DeepSeek’s lower prices come with a different risk profile from self-hosted models, and that trade-off is at the core of every procurement review. When enterprises run open-source models on their own infrastructure, prompts and outputs stay inside existing security controls. By contrast, direct payments to DeepSeek’s hosted service send and receive data through external servers, raising data residency, compliance, and vendor-risk questions. Ramp’s lead economist, Ara Kharazian, emphasized this point, noting that “this is not just self-hosted open source usage. Firms are sending and receiving data through DeepSeek directly.” Security and compliance teams now have to weigh the savings from cheaper AI solutions against exposure created by sending sensitive or regulated information to an additional provider. Many are responding with stricter scoping, limiting early trials to low-sensitivity workloads before they even consider broad deployment.
What DeepSeek’s Momentum Signals for the AI Market
DeepSeek’s position at the top of Ramp’s June trending list signals a broader market shift: pricing and efficiency are starting to compete with vendor lock-in. While DeepSeek’s adoption remains small in Ramp’s index, its momentum shows that procurement teams are more willing to test an AI alternative when budget pressure mounts. The same list includes other lower-cost infrastructure providers, such as Fireworks AI, fal AI, DeepInfra, and Vast.ai, reinforcing the sense that buyers are building a more mixed stack instead of relying on a single premium vendor. This does not yet mean a collapse in demand for leading models. Instead, it suggests a growing two-tier market, where high-end systems handle the most demanding use cases and cost-effective AI models cover routine tasks. Over time, that mix could weaken the grip of any single provider and give enterprises more flexibility to move workloads as pricing and performance change.
