Gemma 3 27B hardware requirements
27.4B parameters. Needs about 21.4 GB at Q4_K_M with 8K context. Model card.
Memory needed at each quantization
| Quant | Quality | 8K context | 32K context |
|---|---|---|---|
| FP16 | Full quality | 59.6 GB | 72 GB |
| Q8_0 | Near-lossless | 33.9 GB | 46.4 GB |
| Q6_K | Excellent | 27.2 GB | 39.7 GB |
| Q5_K_M | Very good | 24.2 GB | 36.7 GB |
| Q4_K_M | Good, most popular | 21.4 GB | 33.9 GB |
| Q3_K_M | Noticeably degraded | 18.2 GB | 30.6 GB |
| Q2_K | Heavily degraded | 15 GB | 27.5 GB |
Run it with Ollama: ollama run gemma3:27b (downloads Q4_K_M by default).
Run it in the cloud
Don't want to buy hardware? Run Gemma 3 27B 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 ~21.4 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 3070 | No | Needs ~21.4 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 3080 10GB | No | Needs ~21.4 GB at Q4_K_M; 10 GB available |
| NVIDIA RTX 3090 | Yes | Q4_K_M · ~21.4 GB · ~34 tok/s |
| NVIDIA RTX 4060 | No | Needs ~21.4 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 4060 Ti 16GB | Yes, with heavy quantization | Q2_K · ~15 GB · ~17 tok/s |
| NVIDIA RTX 5060 Ti 16GB | Yes, with heavy quantization | Q2_K · ~15 GB · ~26 tok/s |
| NVIDIA RTX 4070 | No | Needs ~21.4 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 4070 Ti Super | Yes, with heavy quantization | Q2_K · ~15 GB · ~39 tok/s |
| NVIDIA RTX 4080 | Yes, with heavy quantization | Q2_K · ~15 GB · ~42 tok/s |
| NVIDIA RTX 4090 | Yes | Q4_K_M · ~21.4 GB · ~36 tok/s |
| NVIDIA RTX 5070 | No | Needs ~21.4 GB at Q4_K_M; 12 GB available |
| NVIDIA RTX 5070 Ti | Yes, with heavy quantization | Q2_K · ~15 GB · ~52 tok/s |
| NVIDIA RTX 5080 | Yes, with heavy quantization | Q2_K · ~15 GB · ~56 tok/s |
| NVIDIA RTX 5090 | Yes | Q6_K · ~27.2 GB · ~48 tok/s |
| AMD Radeon RX 7800 XT | Yes, with heavy quantization | Q2_K · ~15 GB · ~36 tok/s |
| AMD Radeon RX 7900 XTX | Yes | Q4_K_M · ~21.4 GB · ~35 tok/s |
| Apple M4 (16GB) | No | Needs ~21.4 GB at Q4_K_M; 10.7 GB available |
| Apple M1 (16GB) | No | Needs ~21.4 GB at Q4_K_M; 10.7 GB available |
| Apple M2 Pro (32GB) | Yes | Q4_K_M · ~21.4 GB · ~7 tok/s |
| Apple M3 Max (64GB) | Yes | Q8_0 · ~33.9 GB · ~8 tok/s |
| Apple M4 Pro (48GB) | Yes | Q8_0 · ~33.9 GB · ~6 tok/s |
| Apple M4 Max (128GB) | Yes | FP16 · ~59.6 GB · ~6 tok/s |
| Apple M2 Ultra (192GB) | Yes | FP16 · ~59.6 GB · ~9 tok/s |