Explore AI
NVIDIA GPU with 24 GB VRAM
The enthusiast tier. 30B-class models at quality quantisations.
The enthusiast tier. 30B-class models at quality quantisations.
- Platform
- NVIDIA
- Memory
- 24 GB vram
A used 24 GB card is the best value in local AI if you are willing to buy second-hand. It runs the largest models that are genuinely useful.
Runs comfortably
- Qwen3 32B 24 GB at Q4 Apache-2.0
- Gemma 4 31B 24 GB at Q4 Apache-2.0
- Qwen3 Coder 30B A3B 20 GB at Q4 · 30B total, 3B active Apache-2.0
- GPT-OSS 20B 16 GB at Q4 · 20B total, 3.6B active Apache-2.0
Will not fit
- GPT-OSS 120Bneeds about 64 GB
Worth knowing
24 GB is where local models start to compete with hosted models for coding, on quality — if not always on convenience.