
NVIDIA / Ada Lovelace
RTX 4090
| Provider / plan | Price / GPU-hour | Rental type | Terms & configuration | Last checked | Provider pricing link |
|---|---|---|---|---|---|
| SaladCloudCommunity / batch | $0.16Updated 2026-11-22 · 15:43 UTCLowest flexible rateCPU/RAM included | ServerlessActive GPU time | 1 GPU minimumConfirm with providerView detailsPublic USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.
|
2026-11-22Auto-collected | View pricing ↗ |
| SaladCloudCommunity / low | $0.2530Updated 2026-11-22 · 15:43 UTCCPU/RAM included | ServerlessActive GPU time | 1 GPU minimumConfirm with providerView detailsPublic USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.
|
2026-11-22Auto-collected | View pricing ↗ |
| Runpodbest-on-demand | $0.34Updated 2026-11-22 · 15:43 UTC | On-demandNo reservation | 1 GPU minimumSee providerView detailsCommunity Cloud: lowest public on-demand GPU-hour list rate among enabled clouds. Not a capacity guarantee; storage/networking extra.
|
2026-11-22Auto-collected | View pricing ↗ |
| SaladCloudCommunity / medium | $0.3470Updated 2026-11-22 · 15:43 UTCCPU/RAM included | ServerlessActive GPU time | 1 GPU minimumConfirm with providerView detailsPublic USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.
|
2026-11-22Auto-collected | View pricing ↗ |
| Vast.aiBest on-demand | $0.3563Updated 2026-11-22 · 15:43 UTC | On-demandNo reservation | 1 GPU minimumCalifornia, USView detailsObserved marketplace offer, not a provider-wide list rate. Total hourly quote for one GPU with the API default disk allocation; bandwidth and additional storage can cost extra. One cheapest verified available on-demand offer per GPU is selected; availability can change.
|
2026-11-22Auto-collected | View pricing ↗ |
| CloudRiftBest on-demand | $0.39Updated 2026-11-22 · 15:43 UTC | On-demandNo reservation | 1 GPU minimumap-east-tw-kn-3noneView detailsAPI list rate; whole-instance charge retained separately. Availability is a point-in-time observation; reservations are not provided by this endpoint.
|
2026-11-22Auto-collected | View pricing ↗ |
| SaladCloudCommunity / high | $0.44Updated 2026-11-22 · 15:43 UTCCPU/RAM included | ServerlessActive GPU time | 1 GPU minimumConfirm with providerView detailsPublic USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.
|
2026-11-22Auto-collected | View pricing ↗ |
Prices are dated listings, not live availability. Compare minimum GPU counts and commitments. Storage, networking and taxes may add charges; CPU/RAM inclusion depends on the provider and is shown beside each serverless rate.
RTX 4090 combines a hardware profile with dated rental offers on xpu live. RTX 4090 is a 24 GB Ada GeForce GPU for graphics and single-GPU compute. It is a candidate for image generation, quantized inference and smaller fine-tuning experiments. Its main sizing constraint is VRAM, especially for long context or concurrent users.
RTX 4090 specifications and configuration
The catalog records Ada Lovelace architecture, 24 GB of VRAM, GDDR6X memory, PCIe form factor, 1008 GB/s memory bandwidth, 450 W maximum board power. These values describe the documented hardware edition. They do not describe the CPU, system RAM or storage of every cloud machine. Open the Specifications tab for the full field-by-field comparison.
24 GB desktop RTX 4090 reference specifications. Laptop GPUs and RTX 4090 D are different products. Consult the manufacturer specification source when checking the exact edition.
Choosing RTX 4090 for your workload
RTX 4090 is worth comparing for Image generation and GPU rendering; 7B–14B-class quantized inference experiments; Adapter fine-tuning with reduced batch sizes. Its documented strengths include Ada Tensor and RT technology; 24 GB on a single card; AV1 encoding support. Match those features to the application you intend to run rather than treating a model name as a performance guarantee.
Before deployment, consider No NVLink; 450 W reference graphics power; Full-precision large-model weights can exceed VRAM. Start with an 8B-class model in a supported 4-bit format. FP16 8B weights alone are about 16 GB before runtime memory; reserve room for KV cache and activations. Adapter fine-tuning needs a separate training memory estimate.
RTX 4090 rental prices and total cost
The rental table separates provider, plan, hourly rate and rental type. Compare on-demand capacity with reserved, spot and serverless terms where those options are listed. The lowest displayed number is not necessarily the lowest total bill for your workload.
Check the last-checked date, minimum GPU count, included CPU and RAM, disk charges and network fees. Confirm availability directly with the provider before paying. Marketplace quotes describe individual machines and can disappear; they are not a permanent provider-wide price.
Compare alternatives before renting
Use the GPU explorer to compare memory and architecture, or browse GPU cloud providers. For RTX 4090, test the same model, precision, batch size and context length on the proposed instance. Record latency, throughput and total cost so your comparison reflects the actual task.
CHOOSING THIS GPU
Is it right for your workload?
24 GB desktop RTX 4090 reference specifications. Laptop GPUs and RTX 4090 D are different products.
Best suited to
- Image generation and GPU rendering
- 7B–14B-class quantized inference experiments
- Adapter fine-tuning with reduced batch sizes
Strengths
- Ada Tensor and RT technology
- 24 GB on a single card
- AV1 encoding support
Things to consider
- No NVLink
- 450 W reference graphics power
- Full-precision large-model weights can exceed VRAM
HARDWARE DETAILS
Inside the GPU
24 GB desktop RTX 4090 reference specifications. Laptop GPUs and RTX 4090 D are different products.
Image: PNY partner card; rental provider chassis and cooler may differ.
- Manufacturer
- NVIDIA
- Architecture
- Ada Lovelace
- VRAM (GB)
- 24
- Memory type
- GDDR6X
- Form factor
- PCIe
- Memory bandwidth (GB/s)
- 1008
- Maximum board power (W)
- 450
- Introduction year
- 2022
- CUDA cores
- 16384
- Tensor cores
- 512 / 4th generation
- Ray tracing cores
- 128 / 3rd generation
- CUDA capability
- 8.9
- Memory ECC
- Not supported on this GeForce reference
- Host interface
- PCIe Gen4
- GPU interconnect
- Not supported
- Hardware partitioning (MIG)
- Not supported by this GPU; do not confuse container sharing with MIG
- Cooling
- Active; board design varies
Hardware capabilities are not a guarantee of access in a cloud instance. Check the exact edition, GPU allocation and server topology in the rental offer.
MODEL MEMORY
Plan your model size
Start with an 8B-class model in a supported 4-bit format. FP16 8B weights alone are about 16 GB before runtime memory; reserve room for KV cache and activations. Adapter fine-tuning needs a separate training memory estimate.
| Example model class | Weight format | Raw weights ≈ | Minimum GPUs for weights only |
|---|---|---|---|
| Llama 3.1 8B ↗ | 4-bit | 4 GB | 1Runtime needs more memory |
| Llama 3.1 8B ↗ | FP16 | 16 GB | 1Runtime needs more memory |
| Qwen2.5 14B ↗ | 4-bit | 7 GB | 1Runtime needs more memory |
| Qwen2.5 32B ↗ | 4-bit | 16 GB | 1Runtime needs more memory |
| Llama 3.1 70B ↗ | 4-bit | 35 GB | 2Runtime needs more memory |
| Llama 3.1 70B ↗ | 8-bit | 70 GB | 3Runtime needs more memory |
| Llama 3.1 70B ↗ | FP16 | 140 GB | 6Runtime needs more memory |
Arithmetic lower bound: rounded parameter count × bits per weight ÷ 8, in decimal GB. Excludes quantization metadata, KV cache, activations, CUDA workspaces and training states. Actual model sizes differ from their rounded names. No context length, batch size or concurrency is guaranteed. GPU counts assume supported model sharding and can be higher in practice. Quantized checkpoints and compatible kernels are required for 4-bit/8-bit execution.
SOFTWARE & PERFORMANCE
Before you deploy
Software compatibility
Use an NVIDIA CUDA-enabled framework/container compatible with this GPU and the host driver. Check PyTorch build and inference-engine kernel requirements before deployment. vLLM documents NVIDIA compute capability 7.5+ as a baseline; support for each model and quantization method still needs checking.
Measured benchmarks
A reproducible, comparable application benchmark has not yet been recorded for this edition. Peak TFLOPS above are not measured tokens per second or image-generation speed.
Compare tests using the same model, precision, GPU count, engine version, input/output length and batch or concurrency. A provider’s server configuration can change the result.
COMPARE YOUR OPTIONS
Alternatives to consider
BUDGET YOUR RUN
GPU cost estimator
GPU compute only: hourly rate × billable hours × GPU count. CPU, RAM, storage, egress, taxes, billing increments and minimum terms may add charges. Serverless hours mean active GPU task time; spot capacity can be interrupted. Reserved commitments and quote-only offers are excluded from this simple hourly estimator.



