Live cloud GPU market · catalog specs/909 GPUs, 313 AI models, and 24 cloud providers
Live AI hardware market
Track live cloud GPU instance prices across providers. Catalog specs and fit estimates when you need them.
Live cloud market
Open cloud pricing →| Family | VRAM | $/hr |
|---|---|---|
| NVIDIA H100 80GB | 80GB | $2.49 |
| NVIDIA A100 40GB | 40GB | $0.89 |
| NVIDIA RTX 4090 | 24GB | $0.40 |
| NVIDIA RTX A6000 | 48GB | $0.80 |
| NVIDIA L40S | 48GB | $1.25 |
| NVIDIA A10 24GB | 24GB | $1.00 |
Cheapest-per-family from the feed — not a ranked “best” list. Full tracker →
Supporting tool
Pick a GPU and a model
Default Q4_K_M · 4k · batch 1. Estimate, not a bench.
Select a GPU and model to check compatibility instantly
Popular checks
Popular compatibility checks
All checksRTX 4090 + Llama 3.1 70B
42.6GB needed (estimate)24GB available
RTX 4070 Ti + Mistral 7B
4.3GB needed (estimate)12GB available
RTX 4090 + Qwen 2.5 72B
43.8GB needed (estimate)24GB available
GeForce RTX 3060 12 GB GA104 + Phi-4
8.5GB needed (estimate)12GB available
RTX 4060 + Llama 3.1 8B
4.9GB needed (estimate)8GB available
Tools
Market first. Fit is supporting.
Cloud GPU PricingCatalog tracks 24 cloud GPU providers. Live offering counts come from the feed on the tracker.Live marketGPU ComparisonSide-by-side specs, benchmarks, and pricing for consumer through enterprise GPUs.909VRAM CalculatorEstimate GPU memory requirements at every quantization level for any model.313 modelsInference CalculatorSelf-hosted vs API: cost-per-token, throughput, and break-even analysis.API vs localModel FinderSearch 313 AI models by capability, size, license, and hardware requirements.313 modelsCompatibility MatrixFull GPU × Model compatibility matrix with quantization-aware VRAM estimates.Matrix view
Weekly brief
Cloud price moves and catalog adds. One email, Monday.