Best LLMs for 48 GB
Pick your memory. We rank every catalogued model that fits at its recommended quantization, by task, and show what the next tier up unlocks.
Written by Jakub Rusinowski · Last updated
Memory tier48 GBFor example RTX A6000, RTX 6000 Ada Generation, L40S
Models that fit138entirely in memory
Best at 48 GB, by task
Score is the task-specific quality score on a 0–100 scale; speed is estimated for the RTX A6000.
The strongest that fit
Top 15, by logic and coding
| Model | Memory · marker = 48 GB | VRAM | Speed on RTX A6000 (est.) | Score |
|---|---|---|---|---|
| Qwen 2.5 72B Instruct Qwen 2.5 Family · Q4_K_M | 43.5 GB | ~10 tok/s | 98 | |
| Qwen 2.5 Coder 32B Qwen 2.5 Family · Q4_K_M | 19.3 GB | ~22 tok/s | 97 | |
| Llama 3.3 70B Instruct Llama 3.3 · Q2_K_XS (Tight) | 20.2 GB | ~21 tok/s | 97 | |
| Qwen 3 32B Qwen 3 · Q4_K_M | 19.8 GB | ~22 tok/s | 97 | |
| DeepSeek R1 Distill Qwen 32B DeepSeek R1 · Q4_K_M | 19.3 GB | ~22 tok/s | 96 | |
| Qwen 2.5 14B Instruct Qwen 2.5 Family · Q4_K_M | 8.5 GB | ~46 tok/s | 95 | |
| Nemotron 70B Instruct Nemotron 70B · Q4_K_M | 42.6 GB | ~11 tok/s | 94 | |
| Qwen 3 14B Qwen 3 · Q4_K_M | 8.9 GB | ~45 tok/s | 94 | |
| Codestral 22B Codestral · Q4_K_M | 13.4 GB | ~31 tok/s | 93 | |
| Qwen 3.6 35B-A3B Qwen 3.6 · Q4_K_M | 21.1 GB | ~158 tok/s | 93 | |
| DeepSeek R1 Distill Qwen 14B DeepSeek R1 · Q4_K_M | 8.5 GB | ~47 tok/s | 92 | |
| Phi-4 (14B) Phi-4 Family · Q4_K_M | 8.5 GB | ~46 tok/s | 92 | |
| Gemma 4 31B Gemma 4 · Q4_K_M | 18.7 GB | ~22 tok/s | 92 | |
| Qwen 2.5 VL 72B Instruct Qwen 2.5 VL · Q4_K_M | 44.3 GB | ~10 tok/s | 92 | |
| Gemma 3 27B Instruct Gemma 3 · Q4_K_M | 16.3 GB | ~26 tok/s | 92 |
What 64 GB unlocks
The strongest models that need the next tier up.
FAQ
What LLMs run well with 48 GB VRAM?
Qwen 2.5 Family, Qwen 2.5 Family, Llama 3.3, Qwen 3, DeepSeek R1 all fit in 48 GB VRAM.
Which GPUs have 48 GB VRAM?
NVIDIA RTX A6000, NVIDIA RTX 6000 Ada Generation, NVIDIA L40S, NVIDIA RTX PRO 5000 Blackwell (48 GB).
All tiers at a glance
| Memory | Models fit | Best overall | Largest that fits |
|---|---|---|---|
| 4 GB | 32 | Gemma 3 4B Instruct | Gemma 3 4B Instruct · 2.4 GB |
| 6 GB | 56 | Granite 3.0 8B Instruct | Qwen 2.5 VL 7B Instruct · 5 GB |
| 8 GB | 71 | Granite 3.0 8B Instruct | Falcon 3 10B Instruct · 6.2 GB |
| 10 GB | 83 | Qwen 2.5 14B Instruct | Qwen 3 14B · 8.9 GB |
| 12 GB | 85 | Qwen 2.5 14B Instruct | StarCoder 2 15B · 9.4 GB |
| 16 GB | 95 | Qwen 2.5 14B Instruct | Mistral Small 3 (24B) · 14.5 GB |
| 20 GB | 111 | Qwen 2.5 14B Instruct | Gemma 4 31B · 18.7 GB |
| 24 GB | 131 | Qwen 2.5 Coder 32B | Qwen 3.6 35B-A3B · 21.1 GB |
| 32 GB | 132 | Qwen 2.5 Coder 32B | Qwen 3.6 35B-A3B · 21.1 GB |
| 48 GB | 138 | Qwen 2.5 72B Instruct | Qwen 2.5 VL 72B Instruct · 44.3 GB |
| 64 GB | 143 | Qwen 2.5 72B Instruct | Qwen3-Coder-Next (80B-A3B MoE) · 48.3 GB |
| 80 GB | 152 | Qwen 2.5 72B Instruct | Qwen 3.5 122B-A10B · 73.7 GB |
Buy this hardware NVIDIA DGX Spark (128GB) — 128 GB VRAM · 150 W board powerDeploy in the cloud now NVIDIA A40 on RunPod — from $0.44/hr · rate checked 2026-08
As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.