IBM signs multi-year agreement with Together AI for $240 million Nvidia-powered inference cluster

IBM has committed to a multi-year, $240 million agreement with Together AI to build a dedicated Nvidia inference cluster on IBM Cloud, the two companies announced on Tuesday.

The new deal hands the open-source AI provider a block of enterprise-grade GPU capacity as it pushes deeper into large corporate accounts.

HGX B300 systems to hit IBM Cloud in early 2027

According to the signed agreement, IBM is set to place a large cluster of Nvidia HGX B300 systems inside IBM Cloud, with the hardware expected to come online in the first quarter of 2027, according to IBM’s newsroom release.

Together AI will use this capacity to serve inference on open-source models. IBM has claimed that it is the first dedicated cluster of this size built for inference on IBM Cloud around the B300, wired with Nvidia’s Spectrum-X Ethernet networking.

Nvidia stated that the configuration can generate 30 times more “AI factory output” than its previous generation, although this claim is not independently verified.

Together AI’s 400 trillion monthly tokens to run on IBM

Together AI had an $800 million Series C funding raise in recent months at an $8.3 billion valuation. This was aimed at expanding what it calls its AI Native Cloud.

The AI company, founded in 2022, rents cloud infrastructure to companies that want to build on open, modular AI stacks instead of closed models. IBM’s press release also mentioned that the company now channels 400 trillion tokens a month through its inference product and helps to power more than a million developers.

The IBM cluster gives Together AI’s planned expansion a platform to run on. CEO Vipul Ved Prakash said the deal was a way to reach more customers without forcing them to pay frontier-model prices.

“Enterprises want the performance of the best frontier models without the closed-model price tag, and that only works if the infrastructure underneath is fast and reliable at scale,” he said in the release. He also added that the cluster is “a big step in our push to make open-source AI the obvious choice for enterprises.”

IBM furthers Nvidia partnership to court AI startups

For IBM, the deal with Together AI is another entry in an organically increasing partnership with Nvidia. Alan Peacock, general manager of IBM Cloud, explained that the deployment was tied to corporate demand for agentic AI, saying IBM and Nvidia are “delivering scalable, economical, enterprise-grade AI infrastructure” to help Together AI move faster.

Dion Harris, a senior director for HPC and AI infrastructure at Nvidia, described AI factories as infrastructure that is becoming as basic as electricity or telecommunications.

IBM stated that the two companies have also been working together on GPU-native data analytics in addition to unstructured data extraction. The companies are also focused on consulting and positioning the Together AI cluster as an important part of a larger push to sell AI capacity to both large enterprises and startups.

The smartest crypto minds already read our newsletter. Want in? Join them.