Which AI models can the NVIDIA RTX 3080 10GB run?
- 10 GB VRAM
- 760 GB/s bandwidth
- NVIDIA
Every model we track, checked against the NVIDIA RTX 3080 10GB at 8K context.
| Model | Runs? | Best fit |
|---|---|---|
| Llama 3.2 1B | Yes | FP16 · ~3.3 GB · ~184 tok/s |
| Llama 3.2 3B | Yes | FP16 · ~8 GB · ~71 tok/s |
| Llama 3.1 8B | Yes | Q6_K · ~8.3 GB · ~69 tok/s |
| Llama 3.1 70B | No | Needs ~46.1 GB at Q4_K_M; 10 GB available |
| Qwen2.5 7B | Yes | Q8_0 · ~9.2 GB · ~56 tok/s |
| Qwen2.5 14B | Yes, with heavy quantization | Q3_K_M · ~9.4 GB · ~63 tok/s |
| Qwen2.5 32B | No | Needs ~22.6 GB at Q4_K_M; 10 GB available |
| Qwen2.5 72B | No | Needs ~47.4 GB at Q4_K_M; 10 GB available |
| Qwen3 8B | Yes | Q6_K · ~8.5 GB · ~68 tok/s |
| Qwen3 30B-A3B | No | Needs ~19.9 GB at Q4_K_M; 10 GB available |
| DeepSeek R1 Distill Qwen 32B | No | Needs ~22.6 GB at Q4_K_M; 10 GB available |
| Mistral 7B v0.3 | Yes | Q8_0 · ~9.4 GB · ~59 tok/s |
| Mistral Nemo 12B | Yes | Q4_K_M · ~9.3 GB · ~62 tok/s |
| Mistral Small 24B | No | Needs ~16.2 GB at Q4_K_M; 10 GB available |
| Mixtral 8x7B | No | Needs ~30 GB at Q4_K_M; 10 GB available |
| Gemma 2 9B | Yes | Q5_K_M · ~10 GB · ~69 tok/s |
| Gemma 2 27B | No | Needs ~20.2 GB at Q4_K_M; 10 GB available |
| Gemma 3 4B | Yes | Q8_0 · ~6.3 GB · ~100 tok/s |
| Gemma 3 12B | Yes, with heavy quantization | Q3_K_M · ~9.8 GB · ~76 tok/s |
| Gemma 3 27B | No | Needs ~21.4 GB at Q4_K_M; 10 GB available |
| Phi-3.5 Mini 3.8B | Yes | Q8_0 · ~7.9 GB · ~112 tok/s |
| Phi-4 14B | Yes, with heavy quantization | Q3_K_M · ~9.5 GB · ~63 tok/s |
| Qwen3 4B | Yes | FP16 · ~9.8 GB · ~57 tok/s |
| Qwen3 14B | Yes, with heavy quantization | Q3_K_M · ~9.2 GB · ~63 tok/s |
| Qwen3 32B | No | Needs ~22.6 GB at Q4_K_M; 10 GB available |
| Qwen3 Coder 30B-A3B | No | Needs ~19.9 GB at Q4_K_M; 10 GB available |
| gpt-oss 20B | Yes, with heavy quantization | Q2_K · ~8.8 GB · ~338 tok/s |
| gpt-oss 120B | No | Needs ~72 GB at Q4_K_M; 10 GB available |
| Llama 4 Scout 17B-16E | No | Needs ~68.3 GB at Q4_K_M; 10 GB available |
Run it in the cloud
Need more than 10 GB? Rent a bigger GPU by the hour.
Rent a cloud GPU by the hour. No upfront cost, cancel any time.