Cassava Technologies Expands AI Factory With NVIDIA Nemotron Models
By making open models available as APIs on Cassava’s AI Factory, CAIMEx gives enterprises greater flexibility to choose the right mix of models for their applications without having to build and operate their own model-serving infrastructure.
Cassava Technologies, a global technology leader, has announced the expansion of its AI Factory capabilities with NVIDIA Nemotron open models, helping enterprises and service providers to grow and innovate with AI more effectively and affordably.
Cassava is deploying NVIDIA Nemotron models as a locally hosted, managed inference service through the Cassava AI Multi-Model Exchange (CAIMEx). By making open models available as APIs on Cassava’s AI Factory, CAIMEx gives enterprises greater flexibility to choose the right mix of models for their applications without having to build and operate their own model-serving infrastructure.
“Expanding our AI Factory across our markets is about creating the capacity and economics needed for businesses to build and deploy AI solutions at scale. By bringing NVIDIA Nemotron into CAIMEx, improving the efficiency of every GPU cycle and introducing flexible token offerings, we are giving enterprises and service providers more choice in how they access and deliver these capabilities to their customers.”
– Ahmed El Beheiry, Group Chief Technology & AI Officer, Cassava Technologies
NVIDIA Nemotron brings efficient reasoning and tool-use capabilities to CAIMEx, helping customers build AI agents that can work through complex tasks and interact with business tools. Its open model weights, training data and recipes also give developers a customizable foundation that they can fine-tune to build specialised agents tailored to their applications.
As businesses mature and the AI agents take on longer-running tasks and access business systems, security and guardrails become critical requirements. These need to be built into the infrastructure, and Cassava is using NVIDIA OpenShell to bring model compute, governed agent execution and independent security enforcement together for customers’ agentic workloads to its AI Factory.
The CAIMEx Token Optimiser is designed to improve the economics of generative and agentic AI by identifying information that has already been processed and, where appropriate, reusing it rather than processing the same content again. This reduces computing requirements and helps customers use their AI tokens more efficiently. The approach applies to both Cassava-hosted and third-party AI models, helping enterprises reduce unnecessary processing as their AI usage scales.
The expansion is the latest step in Cassava’s development of an integrated AI ecosystem, giving businesses access to locally hosted infrastructure, models, and services, fostering trust and confidence in data security and control while supporting their growth.
