Can GeForce RTX 4070 run Mistral Nemo 12B?

✅ Yes, runs great!

Fits in GPU memory (~41 tok/s)

Memory Needed
10.4 GB
Q5_K_M, 8K context
Est. Speed
~41 tok/s
Generation speed
GPU VRAM
12 GB
GDDR6X

💡 Recommended

For best quality that fits: Q5_K_M quantization

All Quantizations

Q8_0
Runs slowly
Memory
14.8 GB
Speed
6 tok/s
Bits/weight
8.5
Quality
Highest
Q6_K
Runs slowly
Memory
11.9 GB
Speed
7 tok/s
Bits/weight
6.6
Quality
High
Q5_K_M
RECOMMENDED
Runs great
Memory
10.4 GB
Speed
41 tok/s
Bits/weight
5.7
Quality
High
Q4_K_M
Runs great
Memory
9.2 GB
Speed
48 tok/s
Bits/weight
4.85
Quality
Good
Q3_K_M
Runs great
Memory
8.0 GB
Speed
60 tok/s
Bits/weight
3.9
Quality
Lower
Q2_K
Runs great
Memory
6.8 GB
Speed
78 tok/s
Bits/weight
3
Quality
Lower

Run with Ollama

ollama run mistral-nemo-q5_k_m

This will download and run the Q5_K_M quantization

View all GPUs for Mistral Nemo 12BView all models for GeForce RTX 4070