Vast.ai
Marketplace of 20,000+ GPUs with supply/demand pricing; RTX 4090 from $0.43/hr (interruptible ~$0.31/hr), H200 from $3.55/hr, B200 from $4.25/hr; any Docker image.
Vast.ai scores 4.6/10 for free value in Functions. The main constraint: Paid — RTX 4090 from $0.43/hr; H100 SXM from ~$1.65/hr; RTX 5090 from $0.66/hr. Marketplace of 20,000+ GPUs with supply/demand pricing; RTX 4090 from $0.43/hr (interruptible ~$0.31/hr), H200 from $3.55/hr, B200 from $4.25/hr; any Docker image.
Free value score
4.6 / 10Show the math
Key metrics
What you get
Paid — RTX 4090 from $0.43/hr; H100 SXM from ~$1.65/hr; RTX 5090 from $0.66/hr. Marketplace of 20,000+ GPUs with supply/demand pricing; RTX 4090 from $0.43/hr (interruptible ~$0.31/hr), H200 from $3.55/hr, B200 from $4.25/hr; any Docker image.
Among the absolute cheapest GPU compute anywhere - RTX 4090 from ~$0.31-0.43/hr and H100s from ~$1.65/hr; unbeatable raw value if you accept variable
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Sources
-
"RTX 5090 from $0.66/hr ... B200 from $4.25/hr ... H200 from $3.55/hr ... RTX 4090 from $0.43/hr"
https://vast.ai/pricing -
"H100s available from ~$1.65/hour for interruptible instances RTX 4090s available from ~$0.31/hour for interruptible instances"
https://northflank.com/blog/cheapest-cloud-gpu-providers
Alternatives
Same category · similar quota$5/mo unlocks Workers + Containers + Durable Objects + R2 ($0.015/GB-month standard, Class A $4.50/M, Class B $0.36/M) with no egress fees.