a100

3 articles · newest first

Where is the cheapest A100 rental right now, and when is the 80GB version worth it?

On Nov 3, 2026 a 40GB A100 rented from $0.40/hr on Vast.ai and an 80GB one from $0.63/hr (median $1.00), against $1.35 on Hyperstack and $1.79 on Runpod. Where the cheap A100s are, why 80GB costs about twice as much, and the jobs that need it.

Aliteq Team · Updated Nov 3, 2026 · 8 min

Thumbnail: Fine-tune an LLM, rent the smallest card that fits. QLoRA needs 6 GB for 8B and 41 GB for 70B. A card reads RTX 4090, Vast median, $0.47/hr, read Nov 3, 2026, with the Vast.ai and Runpod wordmarks.

Which cloud GPU should you rent to fine-tune an LLM, and what does a run cost?

Pick the card by the VRAM your fine-tune needs, not by the biggest name. On Nov 3, 2026 a 24GB RTX 4090 rented for a $0.47/hr median on Vast.ai ($0.89 on Runpod), enough for QLoRA up to about 14B; a 70B QLoRA needs an 80GB A100 from about $0.95/hr. Here is the card for each job and the math for a run.

Aliteq Team · Updated Nov 3, 2026 · 9 min

Night line-art GPU card with the Vast.ai and Runpod logos and the headline: Llama 70B on one rented 48 GB card. A price card reads RTX A6000 48 GB, Vast on-demand, from $0.35 an hour, read Nov 3, 2026.

What does it cost per hour to run Llama 70B on a rented GPU, and which card?

One 48 GB card holds Llama 3.3 70B at Q4_K_M (a 42.5 GB file), and on Nov 3, 2026 the cheapest was an RTX A6000 on Vast from about $0.35 an hour on-demand, or an A40 or RTX A6000 on Runpod at $0.59. When you need an 80 GB card instead, what an evening really costs, and when a per-token API is cheaper than any GPU.

Aliteq Team · Updated Nov 6, 2026 · 8 min