Which GPU runs which model · Llama
Which GPU runs Llama 3.3 70B, and what it costs per hour
Llama 3.3 70B (70B) takes about 140 GB of VRAM as FP16 weights and about 40 to 43 GB at Q4_K_M, before the KV cache for your context window. The table lists the cards on cloudgpu.app by VRAM class, whether the model fits, and the price you will actually be billed, read on 2026-09-15.
Recommended card: A800 80G (80 GB) at $1.09 per hour, the lowest-priced card in a VRAM class that fits.
| VRAM | Cards | Price per hour | Fit |
|---|---|---|---|
| 24 GB | RTX 3090 24G · RTX 4090D 24G · RTX 4090 24G | $0.21 – $0.33/h | Does not fit does not fit on one card; two cards at Q4 is slow |
| 32 GB | RTX 5090 32G | $0.59/h | Does not fit does not fit |
| 48 GB | RTX A6000 48G · RTX 4090 48G | $0.44 – $0.65/h | Tight fit Q4 fits with about 8k context |
| 80 GB | A800 80G · A100 80G | $1.09 – $1.49/h | Fits Q4 comfortable with 32k context; FP16 needs two cards |
| 96 GB | H20 96G · RTX PRO 6000 96G | $1.29 – $1.49/h | Fits Q4 comfortable; Q8 tight |
Prices are the retail rates on cloudgpu.app read 2026-09-15 and equal what the billing system charges; VRAM figures were checked on 2026-09-08. Stock changes during the day.
Reading the table
"Fits" is judged for the Q4_K_M weights plus a working context window; FP16 needs roughly twice the VRAM. The note under each verdict says how much context you get.
Data-centre cards (80 GB and 96 GB) are not cheap on cloudgpu.app. If the model fits on a consumer card, the consumer card is the better deal. The prices here are the same numbers the billing system charges.
Questions about running Llama 3.3 70B
How much VRAM does Llama 3.3 70B need?
Llama 3.3 70B has 70B parameters. The FP16 weights take about 140 GB; at Q4_K_M the weights take about 40 to 43 GB. Add 1 to 8 GB for the KV cache depending on context length and batch size.
What is the cheapest GPU that runs Llama 3.3 70B?
On cloudgpu.app the lowest-priced card we recommend for Llama 3.3 70B is the A800 80G with 80 GB at $1.09 per hour (Q4 comfortable with 32k context; FP16 needs two cards). Prices read 2026-09-15; the pricing page shows live stock.
Can I run Llama 3.3 70B on an RTX 4090 with 24 GB?
No. On a 24 GB card: does not fit on one card; two cards at Q4 is slow. You need a card with at least 80 GB; on cloudgpu.app that is the A800 80G at $1.09 per hour.
What does it cost per hour to run Llama 3.3 70B on cloudgpu.app?
From $1.09 per hour on the A800 80G, billed per minute. A 20-minute test is about $0.36; one hour is pre-authorised when the instance starts and unused minutes are refunded on stop. Cards in higher VRAM classes cost more; the table above lists each.
How do I start Llama 3.3 70B on a rented GPU?
Open Rent a GPU, choose the Ollama template and a card from the table above. The instance is ready in about 100 seconds with a public HTTPS endpoint; then run "ollama pull llama3.3:70b" in the terminal, or request model "llama3.3:70b" through the API and Ollama downloads it on first use.
Do I need a credit card or a Chinese account?
No Chinese account, phone number or ID is required. You can pay by card (Visa or Mastercard, any amount from $5) or USDT (TRC-20). A first top-up of $10 or more gets a $5 bonus, so a $10 card payment starts you with $15 of GPU time.