Runpod Cloud GPU Charge | Generated by AI
Question: What company is RUNPOD.IO MOORESTOWN US, the merchant on a USD 20 charge dated 2026-06-29?
Answer:
That charge is from Runpod, a cloud GPU provider for AI workloads. Headquartered in Moorestown, NJ, founded in 2022, developer-first marketplace — which matches the “MOORESTOWN US” descriptor on your statement.
What they do, relevant to your stack:
- Pods — dedicated GPU instances for dev/training/fine-tuning, full control over container, drivers, environment. Spin up in under 30 seconds, billed by the second, no egress fees.
- Serverless — autoscaling inference endpoints, scale to zero, FlashBoot for sub-200ms cold starts.
- Clusters — multi-node distributed training (800–3200 Gbps interconnect).
Pricing context: pricing starts at $0.24/GPU-hour, averaging ~$1.33 across on-demand configs, with 39 GPU types (4090, L40S, A100, H100, B200). Community Cloud (peer providers, ~20–30% cheaper) vs Secure Cloud (enterprise uptime). They raised $100M led by Summit Partners and crossed one million developers in 2026.
The USD 20 charge is almost certainly a credit top-up — Runpod runs on a prepaid balance model, and $20 is the default minimum auto-load/manual top-up amount. If you didn’t initiate it, check your account’s auto-reload setting in billing; if you did, it’s just your prepaid compute balance. Given you run training workloads on AMD MI300X and RTX 4070, this is likely you (or auto-reload) funding a Runpod pod for NVIDIA-side experiments.
If you don’t recognize it at all, the move is: log into runpod.io → Billing → check transaction history and disable auto-reload, then rotate any API keys if you suspect account compromise.
References: