Open-weight AI Models Now Dominate Enterprise Workloads
Enterprises are choosing portable, open models over proprietary APIs to maintain control and reduce long-term vendor lock-in.
Updated on Oct. 27, 2026 in Artificial Intelligence
Live Poll
Would you prioritize using multiple cloud providers to ensure greater control over your AI data?
Enterprises are increasingly shifting toward open-weight AI models to gain greater control over their internal workflows. New data suggests these models now account for more than 80% of token usage, a significant rise from 20% just two years ago.
Why it matters
Organizations are prioritizing AI workload portability to mitigate risks associated with vendor lock-in, regulatory pressure, and high token costs. By leveraging multiplatform services, enterprises aim to maintain sovereignty over their data and infrastructure.
Open-weight models have grown to represent 80% of total token usage, compared to their 20% share in 2024. This trend underscores a broader move toward platform-agnostic architectures that allow workload migration between on-premises systems and sovereign cloud environments.
The players
Cloudera
A provider of enterprise data services that manages AI workloads and infrastructure across on-premises and cloud environments.
The details
Open-weight models—AI architectures with publicly accessible parameters and weights—allow enterprises to host models internally rather than relying on proprietary APIs. Cloudera enables this by providing data management services that optimize GPU capacity across disparate environments, ensuring workloads remain portable. This architecture allows organizations to implement agentic workflows—autonomous AI agents designed to execute multi-step tasks—while maintaining consistent audit trails and security controls.
Timeline
Open-weight models accounted for 20% of token use in 2024.
The Cloudera analyst discussed AI portability trends at EVOLVE26 in October 2026.
The Tech Race
This trend marks a decisive shift away from the centralized cloud model toward distributed, sovereign AI infrastructure. As enterprises move to retain control over their model outputs, the competition has moved from pure model performance to the ability to effectively manage portable AI workloads.
CIOs and IT leaders can expect to increase focus on documentation regarding agent control, regulatory compliance, and audit trails for these models. Infrastructure teams will likely prioritize tools that facilitate model movement between cloud and internal servers to ensure long-term operational flexibility.
The takeaway
The rapid migration to open-weight models signals that enterprise demand for autonomy is currently outpacing the convenience of closed-model APIs. Readers should track upcoming requirements for standardized audit trails as companies push for more rigorous regulatory compliance in AI deployment.
Further reading
For more on the development of open-source architectures, visit our coverage of Artificial Intelligence.
Source note: This article includes information reported by SiliconANGLE.
Live Poll
Would you prioritize using multiple cloud providers to ensure greater control over your AI data?
