Persistent Linux computers for AI agents, from about $0.03 an hour
Stage
Established
Not given
Founded
Not given
Not given
Based in
US
Not given
Pricing model
Not given
Paid
Pricing
Not given
ac-2: $0.000008 per second ($0.0288 per hour), $13.50 monthly cap. ac-4: $0.000016 per second ($0.0576 per hour), $26.10 cap. ac-8: $0.000032 per second ($0.1152 per hour), $49.50 cap. ac-16: $0.000064 per second ($0.2304 per hour), $89.10 cap. ac-32: $0.000128 per second ($0.4608 per hour), $178.20
Key features
GroqMetal: dedicated bare-metal infrastructure tuned for speed and reliability
GroqCore: production-ready inference stack with no infrastructure expertise required
GroqAssured: enterprise-grade governance, auditability, and control
256 LPUs per rack with 40 PB/s SRAM bandwidth
1,000 tokens per second per user throughput
128 GB on-chip SRAM per rack
315 PFLOPS of FP8 inference compute
13 globally distributed data centers across four continents
Full Debian 13 Linux VM with root access
Docker and Docker Compose included
Chromium desktop with screenshots and input actions
Private HTTPS preview URLs for any port
SSH access including VS Code over SSH
Pause and resume with state preservation
HTTPS API with MCP server support
Per-second billing with monthly caps
Snapshots and private images
Spend controls and monthly limits
What makes it different
Fast inference at scale without sacrificing affordability
LPU pioneered technology integrated with NVIDIA GPUs
Fully integrated platform combining infrastructure, inference, and control
Persistent state preserved across pause and resume cycles