IBM Partners with Together AI for $240M Inference Cluster on IBM Cloud

IBM is making a significant investment in enterprise AI inference by entering a $240 million partnership with Together AI. This collaboration aims to deploy a large-scale inference cluster on IBM Cloud, with the complete rollout expected by 2027.

The partnership combines IBM’s robust enterprise cloud infrastructure with Together AI’s GPU-accelerated platform, which is designed for high-performance model training and inference. Together AI claims its technology can deliver inference times up to twice as fast as standard setups, along with notable cost savings.

This strategic move positions IBM Cloud as a premier destination for enterprises looking to run AI models at scale, rather than focusing solely on model development. Together AI, established in 2020, specializes in providing the necessary infrastructure for open-source AI models, enabling organizations to deploy models without the need to build their own GPU clusters.

In May 2026, IBM Cloud launched Red Hat AI Inference services, which include governance features tailored for enterprises utilizing open-source models. This partnership with Together AI further enhances IBM's AI infrastructure strategy, particularly in addressing the needs of enterprise clients regarding inference performance and cost efficiency.

For Together AI, this collaboration represents a significant opportunity to gain access to IBM's extensive enterprise customer base, ensuring revenue stability through 2027. Both companies share a commitment to making open models viable for enterprise applications, aligning their business models for mutual benefit.