Real-time GPU rental pricing from 50+ cloud providers. Track the Yotta Compute Index (YCI), compare pricing, and analyze market trends.
24h change
24h High
$2.63
24h Low
$2.31
7d Change
+5.2%
30d Change
-3.1%
High-End Index
H100, H200, B200
Mid-Range Index
A100, L40S
Consumer Index
RTX 4090, RTX 3090
Spot Market Index
All spot pricing
Average GPU rental price across all providers
Per-GPU Breakdown
Hopper · 80 GB HBM3
Hopper · 141 GB HBM3e
Ampere · 80 GB HBM2e
Ampere · 40 GB HBM2e
Ada Lovelace · 48 GB GDDR6X
Ada Lovelace · 24 GB GDDR6X
Ampere · 24 GB GDDR6X
Blackwell · 192 GB HBM3e
Full Comparison
| Provider | GPU | Price/hr | Type | Region | Min Commit | |
|---|---|---|---|---|---|---|
Vast.ai | RTX 4090 | $0.24 | Spot | Global | None | |
Vast.ai | RTX 4090 | $0.34 | Community | Global | None | |
Yotta | RTX 4090 | $0.38 | On-demand | US/EU/Asia | 1 sec | |
RunPod | RTX 4090 | $0.44 | On-demand | US/EU | None | |
Yotta | L40S 48GB | $0.65 | On-demand | US/EU/Asia | 1 sec | |
Lambda | L40S 48GB | $0.69 | On-demand | US West | 1 hr | |
CoreWeave | L40S 48GB | $0.74 | On-demand | US East | 1 hr | |
Vast.ai | A100 80GB | $0.85 | Spot | Global | None | |
AWS | L40S 48GB | $0.89 | On-demand | Global | 1 hr | |
Yotta | A100 80GB | $1.29 | On-demand | US/EU/Asia | 1 sec | |
RunPod | A100 80GB | $1.42 | On-demand | US/EU | None | |
Lambda | A100 80GB | $1.49 | On-demand | US West | 1 hr | |
CoreWeave | A100 80GB | $1.85 | On-demand | US East | 1 hr | |
Vast.ai | H100 80GB | $1.99 | Spot | Global | None | |
Fluidstack | H100 80GB | $2.19 | On-demand | US/EU | 1 hr | |
AWS (p4d) | A100 80GB | $2.21 | On-demand | Global | 1 hr | |
Yotta | H100 80GB | $2.35 | On-demand | US/EU/Asia | 1 sec | |
Lambda | H100 80GB | $2.49 | On-demand | US West | 1 hr | |
CoreWeave | H100 80GB | $2.85 | On-demand | US East | 1 hr | |
RunPod | H100 80GB | $3.29 | On-demand | US/EU | None | |
Azure | H100 80GB | $3.67 | On-demand | Global | 1 hr | |
Yotta | H200 141GB | $3.69 | On-demand | US/EU/Asia | 1 sec | |
Lambda | H200 141GB | $3.89 | On-demand | US West | 1 hr | |
GCP | H100 80GB | $3.98 | On-demand | Global | 1 min | |
CoreWeave | H200 141GB | $4.10 | On-demand | US East | 1 hr | |
AWS (p5) | H100 80GB | $4.20 | On-demand | Global | 1 hr | |
AWS | H200 141GB | $5.80 | On-demand | Global | 1 hr |
27 results · Prices updated every 5 minutes
Data Sources
AI infrastructure & open-source models
GPUs
8
Avg Price
$1.67
Uptime
99.95%
High-end GPU clusters
GPUs
4
Avg Price
$2.39
Uptime
99.9%
ML-optimized instances
GPUs
4
Avg Price
$2.14
Uptime
99.8%
Spot & community GPUs
GPUs
3
Avg Price
$0.81
Uptime
97.5%
Serverless GPU inference
GPUs
2
Avg Price
$1.72
Uptime
99.5%
Enterprise cloud
GPUs
3
Avg Price
$3.63
Uptime
99.99%
AI/ML platform
GPUs
2
Avg Price
$3.98
Uptime
99.95%
Set custom alerts for any GPU model. Get notified via email or webhook when prices drop below your target threshold.
Free for all registered users
Prices are collected every 5 minutes from 50+ cloud providers via API integrations and web scraping. All prices are normalized to USD per GPU-hour.
The YCI is a volume-weighted average of on-demand prices across all tracked GPUs and providers. Sub-indices are segmented by GPU tier and pricing type.
Outlier detection algorithms filter anomalous prices. Provider availability and uptime are verified independently. Historical data is immutable once published.