Coming Soon
Public Cloud for
Public Good
A world-class AI Cloud Platform for California public agencies, universities, and institutions.
Platform
A Full-Stack Cloud Platform
Compute, storage, and AI services — the same capabilities as the commercial cloud, operated as public infrastructure for California government agencies, universities, and public institutions.
Cloud Storage
Store and manage your organization's data securely in California. Access records, documents, and application data from any system or workflow.
StorageAI Services
Run AI tools on your agency's data without sending it to a commercial cloud provider. Access AI capabilities from your existing applications and deploy tools trained on your own data.
AICustom AI
Build AI tools tailored to your agency's needs, trained on your own data. Run workloads on CalCompute infrastructure — your data never leaves California throughout the process.
ComputeHosted by
10 world-class research universities — home to Turing Award winners, Nobel laureates, and the researchers defining modern AI. When the University of California builds infrastructure, it's built to research standards.
Why CalCompute
Built for California's Public Sector
Commercial clouds are built for private enterprise. CalCompute is built for public institutions — with the data controls, open standards, and procurement model that government workloads require.
No Vendor Lock-In
Your data and tools work with any system — you're never locked into proprietary formats, and switching platforms is always straightforward. Open standards throughout, no exit fees.
California Data Residency
Your data stays in California. For agencies operating under HIPAA, FERPA, CJIS, or other compliance frameworks, keeping sensitive data within state boundaries is non-negotiable — and CalCompute is built around that requirement.
Predictable Public Pricing
Compute is allocated, not metered. Public agencies get flat, predictable costs that fit government procurement and budget cycles — no variable billing, no surprise charges, no per-token pricing.
For Researchers
Academic & Research Access
UC-affiliated researchers and California academic institutions receive free compute allocations. CalCompute is built for the technical demands of modern AI research.
GPU Compute
Access GPU clusters for distributed training runs, batch jobs, and interactive Jupyter sessions. CUDA and ROCm environments pre-configured. Allocations available for both short exploratory experiments and multi-week training jobs.
ComputeResearch Storage
Petabyte-scale object storage with an S3-compatible API. Built for large datasets, model checkpoints, and training artifacts. High-throughput access for data-intensive workloads — no egress fees.
StorageAI Development APIs
Serve and query open-weights models through an OpenAI-compatible REST API. Deploy custom inference endpoints at scale. Allocation-based access — no per-token pricing, no usage caps that interrupt long experiments.
InferenceUpdates