GPU class · 48 GB
RTX 6000 Ada
What does RTX 6000 Ada cost and what fits on it?
RTX 6000 Ada has 48 GB and costs $0.61 per hour to train on, $0.71 to serve. 36 models in the catalogue fit half-precision LoRA on it and 5 more fit in four-bit.
The fastest 48GB class on the rate card. Where the A40 is the value option at this capacity, this is the one to pick when the run is long enough that throughput dominates the bill — a 14B LoRA that finishes in half the time can cost less overall despite the higher rate.
When to pick this class
The fastest 48GB class on the rate card, and the one to pick when the run is long enough that throughput dominates the bill. A 14B LoRA that finishes in half the time can cost less overall than the same job on an A40 despite a rate about forty-five per cent higher.
The arithmetic is worth doing rather than assuming, because it flips with run length. Under an hour, the cheaper class usually wins; over several hours, this one usually does. The cost calculator makes the comparison concrete for your own dataset size.
Rates
| VRAM | 48 GB |
|---|---|
| Training | $0.61 / GPU-hourMetered per GPU-second, one-minute floor |
| Serving | $0.71 / GPU-hourWhile a replica is resident |
| Warm for a day | $17.0424 hours resident, regardless of traffic |
| Warm for 30 days | $511.20Set the endpoint to scale to zero if nobody is waiting |
| Models — LoRA | 36 |
| Models — QLoRA only | 5 |
Largest models that fit for LoRA
Half precision, frozen base, within 48 GB.
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
Phi-4 14B
14BNeeds 28 GB · MIT
Mistral Nemo 12B
12BNeeds 24 GB · Apache 2.0
Gemma 3 12B
12BNeeds 24 GB · Gemma Terms of Use
Falcon 3 10B
10BNeeds 22 GB · TII Falcon LLM
Models that need four-bit training here
These exceed 48 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 RTX 6000 Ada on capacity and rate.
| Class | VRAM | Training | Serving | Versus this one |
|---|---|---|---|---|
| A100 40 GB | 40 GB | $1.17 | $1.35 | 8 GB less, $0.56/hr dearer |
| A40 | 48 GB | $0.42 | $0.48 | same capacity, $0.19/hr cheaper |
| A6000 | 48 GB | $0.65 | $0.75 | same capacity, $0.04/hr dearer |
| L40S | 48 GB | $0.78 | $0.90 | same capacity, $0.17/hr dearer |
Frequently asked questions
How much does RTX 6000 Ada cost per hour?
$0.61 per GPU-hour for training and $0.71 for serving, metered per GPU-second above a one-minute floor. A continuously warm endpoint on this class is about $17.04 a day, or $511.20 over thirty days.
What models fit on RTX 6000 Ada?
36 models in the catalogue fit half-precision LoRA training within 48 GB, and 5 more fit in four-bit. 37 can be served from this class before accounting for the attention cache.
Is RTX 6000 Ada 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.