Memory tier · 32 GB
Lowest-priced 32 GB GPU listings
2 GPU models carry 32 GB: V100 and RTX 5090. The lowest-priced listing is $0.24 per GPU-hour for the V100 at Verda. Read 12 November 2026.
- Lowest listed
- $0.24
- Per GB-hour
- $0.008
- Providers
- 10
- One day
- $5.86
V100, Verda
lowest listed over memory
with a listing in the tier
one GPU at the lowest listing
Tier
GPU models with 32 GB
| GPU | Memory | Lowest-priced listing | $/GPU-hr | $/GB-hr | Providers | Source |
|---|---|---|---|---|---|---|
| V100datacenter | 32 GB | Verda | $0.24 | $0.008 | 3 | page |
| RTX 5090consumer | 32 GB | packet.ai | $0.36 | $0.011 | 8 | page |
The tier is the card's stated memory, exactly 32 GB. Listings are provider-set on-demand prices in effect today, per GPU-hour. The rule for the lowest-priced listing is simply the lowest price among them.
Listings
Lowest-priced 32 GB listings, provider by provider
| # | GPU | Provider | Configuration | $/GPU-hr | Source |
|---|---|---|---|---|---|
| 1 | V100 | Verda | 1× · 1x Tesla V100 16GB (4 vCPU, 18 GB RAM) | $0.24 | page |
| 2 | RTX 5090 | packet.ai | 1× · Dynamic (shared card) | $0.36 | page |
| 3 | V100 | Runcrateours, live | 1× | $0.40 | page |
| 4 | RTX 5090 | GPUhub | 1× · 1x RTX 5090 | $0.52 | page |
| 5 | RTX 5090 | Novita AI | 1× · GPU instance, on-demand (sample configuration) | $0.56 | page |
| 6 | RTX 5090 | CloudRift | 1× · On-demand (max 2 GPUs per instance) | $0.65 | page |
| 7 | RTX 5090 | Nova Cloud | 1× · 1x RTX 5090, 88 GB system RAM | $0.66 | page |
| 8 | RTX 5090 | Runcrateours, live | 1× | $0.72 | page |
| 9 | V100 | OVHcloud | 1× · t1-45 | $0.77 | page |
| 10 | RTX 5090 | SynpixCloud | 8× · 8x RTX 5090 | $0.89 | page |
| 11 | RTX 5090 | RunPod | 1× · Secure Cloud RTX 5090 | $1.19 | page |
One line per provider and GPU, its lowest price. The fifteen lowest are shown of 11.
Or more
The lowest-priced ways to get at least 32 GB
| GPU | Memory | Lowest-priced listing | $/GPU-hr | $/GB-hr | Providers | Source |
|---|---|---|---|---|---|---|
| V100datacenter | 32 GB | Verda | $0.24 | $0.008 | 3 | page |
| RTX A6000workstation | 48 GB | Thunder Compute | $0.35 | $0.007 | 7 | page |
| RTX 5090consumer | 32 GB | packet.ai | $0.36 | $0.011 | 8 | page |
| L40Sdatacenter | 48 GB | Runcrateours, live | $0.55 | $0.011 | 14 | page |
| A40datacenter | 48 GB | RunPod | $0.59 | $0.012 | 3 | page |
| RTX PRO 6000workstation | 96 GB | packet.ai | $0.76 | $0.008 | 14 | page |
| RTX 6000 Adaworkstation | 48 GB | Novita AI | $0.77 | $0.016 | 2 | page |
| L40datacenter | 48 GB | Thunder Compute | $0.79 | $0.016 | 6 | page |
Every GPU model with 32 GB or more, ordered by lowest listed price, eight lowest. A bigger card can cost less than the card at exactly 32 GB; the memory is then free, but the card may be slower or older.
Arithmetic
What 32 GB holds
Weights alone, at 2 bytes per parameter (FP16), 32 GB holds about 16B parameters; at 1 byte (FP8), 32B; at half a byte (INT4), 64B. A 14 GB 7B model at FP16 fits in 32 GB, and a 26 GB 13B model at FP16 fits. These figures leave out the KV cache and activations, which take more memory as context and batch size grow, so treat them as a ceiling.
Questions
What is the lowest price for a 32 GB GPU?
$0.24 per GPU-hour for the V100 at Verda, read 2026-11-12.
Which GPUs have 32 GB?
V100 (3 providers) and RTX 5090 (8 providers).
What does a 32 GB GPU cost per GB of memory?
At the lowest listing, $0.008 per GB-hour. The median of provider lows across the tier is $0.30 per GPU-hour.
Other memory tiers
Disclosure
Low-Cost GPU Finder is published by the Runcrate team, who also run Runcrate, an AI cloud. Runcrate is listed in the table, marked "ours, live". It is sorted by price like every other row, with no pinned position. Other providers' prices are read from their own public pricing pages. We have no paid placement.