GPU class · 80 GB
A100 80 GB
What does A100 80 GB cost and what fits on it?
A100 80 GB has 80 GB and costs $1.68 per hour to train on, $1.94 to serve. 40 models in the catalogue fit half-precision LoRA on it and 3 more fit in four-bit.
Eighty gigabytes is the threshold where 70B-class models become fine-tunable in four-bit on a single GPU, and where 32B LoRA stops needing compromises. This is the first class on the rate card that opens the large tier at all.
When to pick this class
Eighty gigabytes is the threshold where 70B-class models become fine-tunable in four-bit on a single GPU, and where 32B LoRA stops needing compromises. This is the first class on the rate card that opens the large tier at all, which makes the step up from 48GB a capability decision rather than a speed one.
Against the H100 at the same capacity, this is about two-thirds of the rate and meaningfully slower. For a long run the H100 often costs less overall; for anything short, the difference in wall time will not repay the difference in rate.
Rates
| VRAM | 80 GB |
|---|---|
| Training | $1.68 / GPU-hourMetered per GPU-second, one-minute floor |
| Serving | $1.94 / GPU-hourWhile a replica is resident |
| Warm for a day | $46.5624 hours resident, regardless of traffic |
| Warm for 30 days | $1396.80Set the endpoint to scale to zero if nobody is waiting |
| Models — LoRA | 40 |
| Models — QLoRA only | 3 |
Largest models that fit for LoRA
Half precision, frozen base, within 80 GB.
Qwen3 32B
32BNeeds 64 GB · Apache 2.0
OLMo 3 32B
32BNeeds 64 GB · Apache 2.0
Qwen3 30B-A3B
30BNeeds 64 GB · Apache 2.0
Gemma 3 27B
27BNeeds 56 GB · Gemma Terms of Use
Mistral Small 3 24B
24BNeeds 48 GB · Apache 2.0
LFM2 24B-A2B
24BNeeds 48 GB · LFM Open
gpt-oss 20B
20.9BNeeds 40 GB · Apache 2.0
Qwen3 14B
14BNeeds 28 GB · Apache 2.0
Models that need four-bit training here
These exceed 80 GB in half precision but fit quantised. Quantised training is slower per step, so compare the total run cost against the next class up rather than the hourly rate alone.
Compared with its neighbours
The classes immediately either side of A100 80 GB on capacity and rate.
| Class | VRAM | Training | Serving | Versus this one |
|---|---|---|---|---|
| A6000 | 48 GB | $0.65 | $0.75 | 32 GB less, $1.03/hr cheaper |
| L40S | 48 GB | $0.78 | $0.90 | 32 GB less, $0.90/hr cheaper |
| H100 80 GB | 80 GB | $2.59 | $2.99 | same capacity, $0.91/hr dearer |
| H200 | 141 GB | $4.55 | $5.25 | 61 GB more, $2.87/hr dearer |
Frequently asked questions
How much does A100 80 GB cost per hour?
$1.68 per GPU-hour for training and $1.94 for serving, metered per GPU-second above a one-minute floor. A continuously warm endpoint on this class is about $46.56 a day, or $1396.80 over thirty days.
What models fit on A100 80 GB?
40 models in the catalogue fit half-precision LoRA training within 80 GB, and 3 more fit in four-bit. 42 can be served from this class before accounting for the attention cache.
Is A100 80 GB the cheapest option?
Cheapest per hour is not the same as cheapest per run. Total cost is the rate multiplied by wall time, so a faster class that finishes sooner often costs less — particularly when the cheaper alternative would require quantised training, which is slower per step.
Last verified 6 August 2026.
Start with the free tier
A magic link creates your account, your tenant and your first API key. No card until you ask for compute.