GPU Cloud Picks by Use Case
Opinionated recommendations across 28 providers, organized by what you're actually trying to do. Each page splits the recommendation three ways — performance, ops, cost — because the cheapest provider is rarely the easiest to operate, and the fastest GPU is rarely the cheapest per token.
Live pricing + 7-day reliability scores update on every page load. Curation refreshed manually — last updated May 2026.
Training
Training picks split three ways. The cost-optimal provider isn't always the performance-optimal one — multi-node training is interconnect-bound, and free egress matters more than the headline $/hr when you're moving terabytes of checkpoints. Here's how three lenses (performance, ops, cost) actually shake out, with live pricing.
View 3 picks →
Inference
Inference economics differ from training: latency, cold-start, billing granularity, and per-token cost all matter more than peak FLOPS. The right GPU for an 8B model is rarely the right GPU for a 405B model. Here's how the three axes split.
View 3 picks →
Fine-tuning
Fine-tuning sits between training and inference: short bursts of GPU work interleaved with eval, dataset shuffling, and human review. Billing granularity (per-second vs per-hour), persistent storage, and dataset egress hit harder than the headline $/hr. Three picks across the axes.
View 3 picks →
Budget
Cheap doesn't mean cheap. A $1.80/hr H100 at 70% reliability costs more in re-runs than a $2.39/hr H100 at 95%. The picks below show the absolute floor, the most reliable budget option, and the ones to actively avoid despite their brand recognition.
View 3 picks →