GPUs with 32 GB VRAM and more
Cards with at least 32 GB of video memory start at €0.46 an hour on VRAMfarm (L40). 32 GB holds a language model of about 48B parameters in 4-bit or 14B in FP16. For images and video: everything that runs on 24 GB, with larger batches and higher resolution.
32 GB on the RTX 5090 gives the speed of the newest consumer card and room that 24 GB lacks: bigger batches, higher resolution, 32B models with longer context.
Cards with 32 GB or more, cheapest first
| GPU | Type | Memory | TFLOPS FP32 | Where | Price per hour | ≈ per month | Action |
|---|---|---|---|---|---|---|---|
| L40 | data centre | 48 GB | ≈90.5 | €0.46/h | €331 per month | Rent | |
| RTX A6000 | workstation | 48 GB | ≈38.7 | €0.72/h | €518 per month | Rent | |
| RTX 5090 | consumer | 32 GB | ≈104.8 | +15 | €0.76/h | €547 per month | Rent |
| L40S | data centre | 48 GB | ≈91.6 | +1 | €1.06/h | €763 per month | Rent |
| A100 | data centre | 80 GB | ≈19.5 | +2 | €1.46/h | €1,051 per month | Rent |
| RTX PRO 6000 | workstation | 96 GB | ≈125 | +8 | €2.12/h | €1,526 per month | Rent |
| H100 | data centre | 80 GB | ≈67 | +1 | €3.90/h | €2,808 per month | Rent |
| H200 | data centre | 141 GB | ≈67 | +2 | €5.43/h | €3,910 per month | Rent |
| B200 | data centre | 192 GB | ≈80 | €16.12/h | €11,606 per month | Rent |
To launch, your balance must cover one hour of the machine. You are billed by the minute for actual use.
What runs on 32 GB
Language models
Up to about 48B parameters in 4-bit (AWQ, GPTQ, GGUF Q4) or 14B in FP16, with room for a normal context. Long context and many parallel users need more. The VRAM calculator works it out for your model.
Images and video
Everything that runs on 24 GB, with larger batches and higher resolution.
Training
QLoRA of models up to about 32B with short sequences, LoRA of 8B with a 16-bit base.
Typical jobs at this size
- Qwen-Image and Wan 2.2 video
- Chat models up to 32B in 4-bit
- LoRA training for Flux and SDXL
Frequently asked questions
What is the cheapest GPU with 32 GB or more?
The NVIDIA L40, from €0.46 an hour on VRAMfarm, billed by the minute.
Which LLM can I run on 32 GB?
About 48B parameters in 4-bit or 14B in FP16, with room for a normal context. For example: Llama 3.1 8B, Qwen3 14B, Qwen3 32B.
Can I combine the memory of several GPUs?
For language models, yes: vLLM and llama.cpp split a model across cards, so 2 × 24 GB can run what needs about 40 GB. For image and video models memory does not add up, one card has to hold the model.
How do I know 32 GB is enough?
Add up the weights (2 GB per billion parameters in FP16, 0.6 GB in 4-bit), the context cache and 10–15% for buffers, or use our VRAM calculator. If the model does not fit, it either fails with “CUDA out of memory” or runs much slower with offloading.
Rent 32 GB of VRAM
Top up from €10, pick a card from the list and start in a few minutes.