NVIDIA RTX PRO 5000 Blackwell 48GB
48 GB decides what fits. 1344 GB/s decides how fast it runs once it does.
Open in the calculator →Best models for 48 GB, on every card that size →
Find one: Amazon ↗ · eBay (new and used) ↗Store links may pay us a commission. They never decide which card is suggested — the memory arithmetic does.
- Largest popular model that fitsKimi-Linear-48B-A3B-Instruct49B params · Q4_K_M · needs 30.4 GB~489 tok/s Faster than you readOpen in the calculator →
- Most downloaded that fitsQwen3.8-27B28B params · Q4_K_M · needs 18.6 GB~62 tok/s Faster than you readOpen in the calculator →
At 8,192 tokens of context with the whole model in device memory. Speeds are estimates from memory bandwidth, not benchmarks run on this card.
226 of 320 fit entirely
226 of the 320 models the engine can size fit entirely in device memory at Q4_K_M where it is published, otherwise the nearest published format, and 8,192 tokens. 18 run with some layers on system memory (32 GB assumed), and 76 do not run at all.
Featured models · Q4_K_M at 8,192 tokens, including offload
| Model | Needs | Verdict | Decode | Calculator |
|---|---|---|---|---|
| Qwen3.5-2B2.3B parameters | 2.32 GB | fits | ~741 tok/s | Open → |
| NVIDIA-Nemotron-3-Nano-4B-BF164.0B parameters | 3.46 GB | fits | ~343 tok/s | Open → |
| Qwen3-VL-8B-Instruct8.8B parameters | 7.04 GB | fits | ~155 tok/s | Open → |
| Qwen3.5-9B9.7B parameters | 7.05 GB | fits | ~188 tok/s | Open → |
| gemma-4-12B-it12B parameters | 9.54 GB | fits | ~112 tok/s | Open → |
| gpt-oss-20b21B parameters | 13.8 GB | fits | ~365 tok/s | Open → |
| Qwen3.8-27B28B parameters | 18.6 GB | fits | ~62 tok/s | Open → |
| Qwen3.6-27B28B parameters | 18.6 GB | fits | ~62 tok/s | Open → |
| Kimi-Linear-48B-A3B-Instruct49B parameters | 30.4 GB | fits | ~489 tok/s | Open → |
| Qwen3.8-Flash-Next180B parameters | 111.6 GB | needs 64.5 GB of system RAM; 32 GB assumed | does not run | Open → |
| DeepSeek-V4-Flash-Vision-Exp305B parameters | 188.1 GB | needs 143.7 GB of system RAM; 32 GB assumed | does not run | Open → |
| GLM-5.3-Flash321B parameters | 198.3 GB | needs 153.4 GB of system RAM; 32 GB assumed | does not run | Open → |
These examples are selected from prominent labs using the catalogue’s latest Hugging Face 30-day downloads and repository-creation freshness signal, with newer releases guaranteed a place. Offloaded rows assume 32 GB of system RAM, and a speed is only shown for a row that runs. Fitting in memory is not the same as loading: whether the runtime and version you have supports each architecture and format on this machine has not been tested here. Each “Open” link carries the same model, format, context and RAM into the calculator.
Fits entirely in NVIDIA RTX PRO 5000 Blackwell 48GB memory — 226 of 320 sized models
The most demanding model that fits is Qwen2.5-VL-72B-Instruct at Q4_K_M: 47.3 GB of the 48 GB, leaving 0.70 GB spare.
Fully resident at 8,192 tokens (or the model’s own maximum, where that is shorter), offload off, at 1344 GB/s. Each row is a run of the engine for this configuration; the rows start with current, prominent releases and “fits” is memory, not a tested runtime. Older or less prominent models remain available through this search and “Show all”.
Showing 40 of 226 models that fit.
| Model | Format | Needs | Spare | Decode | Downloads / 30d | Calculator |
|---|---|---|---|---|---|---|
| Qwen3.8-27BQwen · 28B params | Q4_K_M | 18.6 GB | 29.4 GB | ~62 tok/s | 6.9M | Open → |
| gemma-4-26B-A4B-itGoogle · 26B params | Q4_K_M | 16.8 GB | 31.2 GB | ~343 tok/s | 13M | Open → |
| gemma-4-31B-itGoogle · 31B params | Q4_K_M | 22.2 GB | 25.8 GB | ~44 tok/s | 9.9M | Open → |
| Qwen3.5-9BQwen · 9.7B params | Q4_K_M | 7.05 GB | 41.0 GB | ~188 tok/s | 9M | Open → |
| Qwen3.5-4BQwen · 4.7B params | Q4_K_M | 3.98 GB | 44.0 GB | ~338 tok/s | 7.8M | Open → |
| Qwen3.6-35B-A3BQwen · 36B params | Q4_K_M | 23.1 GB | 24.9 GB | ~493 tok/s | 3.3M | Open → |
| gemma-4-12B-itGoogle · 12B params | Q4_K_M | 9.54 GB | 38.5 GB | ~112 tok/s | 1.9M | Open → |
| Qwen3.6-27BQwen · 28B params | Q4_K_M | 18.6 GB | 29.4 GB | ~62 tok/s | 2.5M | Open → |
| Qwen3.5-2BQwen · 2.3B params | Q4_K_M | 2.32 GB | 45.7 GB | ~741 tok/s | 4.9M | Open → |
| gemma-4-E4B-itGoogle · 8.0B params | Q4_K_M | 5.86 GB | 42.1 GB | ~302 tok/s | 4.4M | Open → |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA · 32B params | Q4_K_M | 20.3 GB | 27.7 GB | ~47 tok/s | 530K | Open → |
| NVIDIA-Nemotron-3-Nano-4B-BF16NVIDIA · 4.0B params | Q4_K_M | 3.46 GB | 44.5 GB | ~343 tok/s | 3.4M | Open → |
| gemma-4-E2B-itGoogle · 5.1B params | Q4_K_M | 4.02 GB | 44.0 GB | ~798 tok/s | 3M | Open → |
| Qwen3.5-0.8BQwen · 873M params | Q4_K_M | 1.46 GB | 46.5 GB | ~1623 tok/s | 2.6M | Open → |
| Qwen3-VL-8B-InstructQwen · 8.8B params | Q4_K_M | 7.04 GB | 41.0 GB | ~155 tok/s | 14.6M | Open → |
| Qwen3.5-27BQwen · 28B params | Q4_K_M | 18.6 GB | 29.4 GB | ~62 tok/s | 1.9M | Open → |
| North-Micro-Vision-InstructCohereLabs · 2.5B params | Q4_K_M | 2.99 GB | 45.0 GB | ~487 tok/s | 180.7K | Open → |
| Qwen3.5-35B-A3BQwen · 36B params | Q4_K_M | 23.1 GB | 24.9 GB | ~493 tok/s | 1.6M | Open → |
| granite-4.2-8bIBM · 8.8B params | Q4_K_M | 7.49 GB | 40.5 GB | ~141 tok/s | 128.4K | Open → |
| GLM-4.7-Flashzai-org · 31B params | Q4_K_M | 20.4 GB | 27.6 GB | ~47 tok/s | 1.8M | Open → |
| granite-4.1-3bIBM · 3.4B params | Q4_K_M | 3.72 GB | 44.3 GB | ~330 tok/s | 521.8K | Open → |
| LFM2.5-2.6BLiquidAI · 2.7B params | Q4_K_M | 2.61 GB | 45.4 GB | ~529 tok/s | 108.1K | Open → |
| Qwen3-VL-4B-InstructQwen · 4.4B params | Q4_K_M | 4.72 GB | 43.3 GB | ~247 tok/s | 3.5M | Open → |
| granite-4.1-30bIBM · 29B params | Q4_K_M | 20.4 GB | 27.6 GB | ~46 tok/s | 301.5K | Open → |
| Qwen3-VL-2B-InstructQwen · 2.1B params | Q4_K_M | 3.05 GB | 45.0 GB | ~447 tok/s | 2.8M | Open → |
| granite-4.2-3bIBM · 3.7B params | Q4_K_M | 3.72 GB | 44.3 GB | ~330 tok/s | 50.6K | Open → |
| LFM2.5-230MLiquidAI · 230M params | Q4_K_M | 1.05 GB | 46.9 GB | ~3689 tok/s | 88.6K | Open → |
| Qwen3-0.6BQwen · 752M params | Q4_K_M | 2.22 GB | 45.8 GB | ~686 tok/s | 29.7M | Open → |
| gpt-oss-20bOpenAI · 21B params | Q4_K_M | 13.8 GB | 34.2 GB | ~365 tok/s | 6.6M | Open → |
| granite-4.2-30bIBM · 29B params | Q4_K_M | 20.7 GB | 27.3 GB | ~46 tok/s | 35.8K | Open → |
| granite-4.1-8bIBM · 8.8B params | Q4_K_M | 7.49 GB | 40.5 GB | ~141 tok/s | 179.3K | Open → |
| NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA · 32B params | Q4_K_M | 20.3 GB | 27.7 GB | ~47 tok/s | 875.2K | Open → |
| LFM2.5-VL-3BLiquidAI · 3.1B params | Q4_K_M | 2.85 GB | 45.1 GB | ~529 tok/s | 26.7K | Open → |
| MiMo-V2.6-Distill-Qwen-9BNEWXiaomiMiMo · 9.4B params | Q4_K_M | 6.90 GB | 41.1 GB | ~188 tok/s | 14.6K | Open → |
| Qwen3-4B-Instruct-2507Qwen · 4.0B params | Q4_K_M | 4.72 GB | 43.3 GB | ~247 tok/s | 3.7M | Open → |
| Ling-3.0-tinyinclusionAI · 7.9B params | Q4_K_M | 5.85 GB | 42.1 GB | ~1066 tok/s | 17.7K | Open → |
| Qwen3-8BQwen · 8.2B params | Q4_K_M | 7.04 GB | 41.0 GB | ~155 tok/s | 10.7M | Open → |
| Olmo-3-7B-Instructallenai · 7.3B params | Q4_K_M | 7.96 GB | 40.0 GB | ~132 tok/s | 481.7K | Open → |
| Qwen3-4BQwen · 4.0B params | Q4_K_M | 4.51 GB | 43.5 GB | ~247 tok/s | 7.8M | Open → |
| LFM2.5-8B-A1BLiquidAI · 8.5B params | Q4_K_M | 6.06 GB | 41.9 GB | ~825 tok/s | 32.4K | Open → |
Sized at the format most people actually download, not at FP16. The catalogue ordered by downloads →
Rent an RTX PRO 5000 by the hour
What the NVIDIA RTX PRO 5000 Blackwell 48GB costs on Vast.ai and RunPod right now, per machine, from one card to eight. Every price is read from the provider's own API.
$0.51per hour
Vast.ai · Marketplace · 98.1% reliableRunPod $0.82/h Community CloudRent on Vast.aiBuy one or rent one?
Your price, your hours, your electricity. Everything else is arithmetic.
Watts start at the NVIDIA RTX PRO 5000 Blackwell 48GB’s published board power (300 W), an upper bound: decoding rarely holds a card at its limit. Rent starts at the cheapest live price (Vast.ai). The electricity price is a placeholder — put yours in.
At 4 h a day, renting costs $744.60 a year.
- That is $62.05 a month, and nothing when the machine is stopped.
- Owning would cost $109.50 a year in electricity, on top of the price.
- Every $1,000 of purchase price takes 1.6 years of this use to earn back.
Enter the price you would pay to see the exact break-even.
On-demand prices for the whole machine. Vast hosts below 98% measured reliability are left out; RunPod’s Community Cloud is vetted third-party hosts and its Secure Cloud is data-centre capacity. Every rentable card compared.
Referral links Vast.ai, RunPod and Novita pay us a share of what you spend if you sign up through these buttons. It costs you nothing, and it never decides an order or a recommendation: both are computed from the live price and the speed, and options that pay us nothing are listed and recommended on the same terms. How we rank
Find the NVIDIA RTX PRO 5000 Blackwell 48GB: Amazon ↗ · eBay (new and used) ↗Store links may pay us a commission. They never decide which card is suggested — the memory arithmetic does.
The RTX PRO 5000 Blackwell 48GB, model by model
One page per model: whether it fits this device, at which formats, and how fast.
The specification behind every figure
What the manufacturer publishes for this device, and the pages it was read from.
Manufacturer specification
| Memory scope | dedicated |
|---|---|
| Capacity | 48 GB |
| Published options | 48 GB |
| Bandwidth | 1344 GB/s |
| Memory type | GDDR7 with ECC |
| Bus width | Not published |
| FP32 peak | Not published |
| Dense matrix peak | Not published without sparsity |
| maximum power consumption | 300 W |
Source ledger
Caveats
- NVIDIA publishes 48 GB and 72 GB parts at the same bandwidth; each is catalogued separately because a discrete card's memory is not configurable.
- NVIDIA publishes no general-purpose peak throughput for this card, so no compute roof is priced for it and time to first token is withheld.
- No dense matrix throughput is catalogued for this device: NVIDIA publishes no dense matrix throughput for this card. Time to first token is withheld rather than priced against a sparsity figure.
Published capacity is a hardware ceiling, not guaranteed free runtime memory. The calculator shows the runtime reserve separately rather than folding it into a single number.
Run the diagnostic on the NVIDIA RTX PRO 5000 Blackwell 48GB →