Llama 3.1 70B hardware requirements
70.6B parameters. Needs about 46.1 GB at Q4_K_M with 8K context. Model card.
Memory needed at each quantization
| Quant | Quality | 8K context | 32K context |
|---|---|---|---|
| FP16 | Full quality | 144.5 GB | 152.5 GB |
| Q8_0 | Near-lossless | 78.3 GB | 86.3 GB |
| Q6_K | Excellent | 61.2 GB | 69.2 GB |
| Q5_K_M | Very good | 53.5 GB | 61.6 GB |
| Q4_K_M | Good, most popular | 46.1 GB | 54.1 GB |
| Q3_K_M | Noticeably degraded | 37.8 GB | 45.8 GB |
| Q2_K | Heavily degraded | 29.8 GB | 37.8 GB |
Run it with Ollama: ollama run llama3.1:70b (downloads Q4_K_M by default).
Run it in the cloud
Don't want to buy hardware? Run Llama 3.1 70B on a rented GPU today.
Rent a cloud GPU by the hour. No upfront cost, cancel any time.
Every GPU and Mac checked
| GPU | Runs? | Best fit |
|---|---|---|
| NVIDIA RTX 3060 12GB | No | Needs ~46.1 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 3070 | No | Needs ~46.1 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 3080 10GB | No | Needs ~46.1 GB at Q4_K_M; 10 GB available |
| NVIDIA RTX 3090 | No | Needs ~46.1 GB at Q4_K_M; 24 GB available |
| NVIDIA RTX 4060 | No | Needs ~46.1 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 4060 Ti 16GB | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 5060 Ti 16GB | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 4070 | No | Needs ~46.1 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 4070 Ti Super | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 4080 | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 4090 | No | Needs ~46.1 GB at Q4_K_M; 24 GB available |
| NVIDIA RTX 5070 | No | Needs ~46.1 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 5070 Ti | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 5080 | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| NVIDIA RTX 5090 | Yes, with heavy quantization | Q2_K · ~29.8 GB · ~41 tok/s |
| AMD Radeon RX 7800 XT | No | Needs ~46.1 GB at Q4_K_M; 16 GB available |
| AMD Radeon RX 7900 XTX | No | Needs ~46.1 GB at Q4_K_M; 24 GB available |
| Apple M4 (16GB) | No | Needs ~46.1 GB at Q4_K_M; 10.7 GB available |
| Apple M1 (16GB) | No | Needs ~46.1 GB at Q4_K_M; 10.7 GB available |
| Apple M2 Pro (32GB) | No | Needs ~46.1 GB at Q4_K_M; 21.4 GB available |
| Apple M3 Max (64GB) | Yes | Q4_K_M · ~46.1 GB · ~6 tok/s |
| Apple M4 Pro (48GB) | Yes, with heavy quantization | Q2_K · ~29.8 GB · ~6 tok/s |
| Apple M4 Max (128GB) | Yes | Q8_0 · ~78.3 GB · ~4 tok/s |
| Apple M2 Ultra (192GB) | Yes | Q8_0 · ~78.3 GB · ~6 tok/s |