GPU

RTX 4080 Ti 16GB for local AI

16GB memory, 678 GB/s bandwidth, and deterministic compatibility links for downloadable local models.

VRAM

16GB

Bandwidth

678 GB/s

Architecture

Ada Lovelace

Memory type

GDDR6X

Models to check on this GPU

• Can it run OpenAI gpt-oss-20b?
• Can it run InternVL3 5-GPT-OSS-20B-A4B-Preview?
• Can it run Kimi-VL-A3B-Instruct?
• Can it run Kimi-VL-A3B-Thinking-2506?
• Can it run Moonlight-16B-A3B-Instruct?
• Can it run DeepSeek Coder V2 Lite Instruct?
• Can it run DeepSeek-V2-Lite?
• Can it run DeepSeek R1 Distill Qwen 14B?