Together AI, Y Combinator Launch Dedicated GPU Cluster to Help AI Startups Overcome Compute Crunch
The initiative gives AI-native startups in YC's portfolio dedicated access to GPU resources for both AI training and inference.
Together AI and startup accelerator Y Combinator (YC) have partnered to launch the first dedicated GPU cluster exclusively for YC-backed startups, to ease one of the biggest challenges facing AI companies today—access to affordable computing infrastructure.
Together AI and @ycombinator are launching the first dedicated YC GPU cluster. Startups in the YC portfolio can now access compute on a few weeks' commitment, instead of a 24-month contract. pic.twitter.com/sXhTjNPKB9
— Together AI (@togethercompute) July 20, 2026
The initiative gives AI-native startups in YC's portfolio dedicated access to GPU resources for both AI training and inference, allowing them to scale applications without committing to expensive long-term compute contracts.
According to Together AI, many early-stage AI companies are now forced to spend a significant portion of their funding on compute infrastructure or raise additional capital solely to secure GPU access.
Under the partnership, YC startups will be able to reserve, provision, and manage GPUs through Together AI's self-service platform, enabling them to deploy compute resources within minutes while retaining full control over usage and billing. The platform supports a range of workloads, from single-node deployments to large-scale AI training as startups grow.
"Excited to be partnering with Y Combinator to offer YC startups broad access to AI infrastructure and services," Vipul Ved Prakash, Together AI Co-founder & CEO, said.
"We’ve been funding more research teams / neolabs than ever, and we’ve built the first YC GPU cluster to power them in partnership with Together AI. If you’re a researcher who needs compute for your startup idea, apply to YC today," Ankit Gupta, Y Combinator General Partner, added.
Beyond infrastructure, founders will also gain access to Together AI's engineering expertise, research, and operational best practices developed through collaborations with leading AI companies.
The dedicated cluster is already operating at full utilisation, though startups can plan and reserve compute capacity months in advance.
The two organisations said they plan to expand the GPU cluster over time as demand grows.
Startups graduating from YC will continue to receive support through Together AI's broader platform, which includes inference, fine-tuning, and model training services, allowing them to scale on the same infrastructure as their businesses mature.
Founded four years ago, Together AI provides cloud infrastructure for the entire generative AI lifecycle and serves more than 8,000 customers, including Cursor, Decagon, Cognition, and ElevenLabs.
