Gradient°

The product is a cloud GPU platform that provides on-demand access to NVIDIA H100 and other GPUs to develop, train, fine-tune, and deploy AI/ML models including large language models and generative media, removing the need to buy and manage own servers. It sells to developers, data teams, and IT teams in startups, SMBs, and enterprises building AI applications, 3D graphics, rendering, and simulation workloads. It is delivered as a SaaS and infrastructure-API platform with notebooks, training machines, scalable API deployments, per-second billing, and managed compute, storage, and networking.

Key features

• Develop AI models in notebooks
• Train and fine-tune models on GPU machines
• Deploy models as scalable API endpoints
• On-demand access to NVIDIA H100 GPUs
• Per-second billing with no commitments
• Pre-configured templates for quick start
• Job scheduling and resource provisioning
• Automatic versioning and tagging
• Scale training with multiple GPU options
• Managed compute, storage and networking
• Cloud hosted virtual desktops
• GPU workflows for ML pipelines