GPU
RTX 4070 SUPER 12GB for local AI
12GB memory, 504 GB/s bandwidth, and deterministic compatibility links for downloadable local models.
- VRAM
- 12GB
- Bandwidth
- 504 GB/s
- Architecture
- Ada Lovelace
- Memory type
- GDDR6X
Models to check on this GPU
- Can it run DeepSeek R1 Distill Qwen 1.5B?
- Can it run Mistral Small 3.1 24B Instruct?
- Can it run DeepSeek R1 Distill Qwen 14B?
- Can it run Qwen2.5-14B-Instruct?
- Can it run Qwen2.5-Coder-14B?
- Can it run Qwen 2.5 Coder 14B Instruct?
- Can it run Qwen 3 14B?
- Can it run Microsoft Phi-4 14B?
- Can it run Phi-4-reasoning-plus?
- Can it run Mistral Ministral 3 14B Instruct?