06/05/2025 - Daily GPU Cloud pricing update - Major price reductions across H100/H200 offerings
Cudo Compute leads today's pricing adjustments with significant reductions:
- 29.2% reduction on H100 SXM single GPU instances
- 14.1% reduction on H100 PCIe instances
- Introduction of more competitive pricing for inference workloads
Hyperstack makes aggressive moves on multi-GPU clusters:
- 20.8% reduction on 8xH100 PCIe bare metal clusters
- New pricing structure for NVLink configurations
- Enhanced memory configurations for ML training workloads
VAST.AI marketplace shows important shifts:
- 15.2% reduction on H200 instances, now at $3.30/gpu/hour
- 21.5% reduction on H100 NVL instances
- Increased availability of A100 and L40S instances
Other notable changes:
- Fluidstack reduces A100 PCIe pricing by 12.7%
- Vultr adjusts 8xH100 cluster pricing with 7.1% reduction
- DataCrunch optimizes L40S pricing with 10.6% reduction
- RunPod introduces competitive MI300X pricing at $2.99/hour
The trend shows aggressive pricing competition especially in the high-end segment (H100/H200), while mid-range options (A100/L40S) are also seeing significant adjustments to remain competitive.
--
You are a GPU Cloud consumer? Sign In to get the latest opportunities pricing:
https://gpucompare.com/
You are a GPU Cloud provider? Sign In to claim your Page and share opportunities:
https://gpucompare.com/providers