Directory

Local AI models

Open-weight models with parameter counts, quantization VRAM, and honest fit against the GPU you already own.

What can I run Grid Table

Qwen3-0.6B 0.8B

2.7/8 GB

10.7/8 GB

Qwen2.5-7B-Instruct 7.6B

6.4/8 GB

Qwen3.6-35B-A3B-NVFP4 19B

13.2/8 GB

Qwen2.5-1.5B-Instruct 1.5B

1.8/8 GB

Qwen2.5-3B-Instruct 3.1B

3/8 GB

OTel-2.0-LLM-31B-IT 31B

22.2/8 GB

gpt-oss-20b 21B

17.1/8 GB

Qwen3.8-Flash-Next 180B

127.6/8 GB

286/8 GB

GLM-5.3 753B

2335.9/8 GB

6/8 GB

Tev1-4B-experimental 4.7B

3.3/8 GB

Tev1-0.8B-experimental 0.9B

0.6/8 GB

AdvancedMathBench-AutoVerifier 36B

25.5/8 GB

gemma-4-E4B 8.0B

5.7/8 GB

Kimi-K3-DSpark 2.2B

36.4/8 GB

d1-omni-600M 0.6B

0.4/8 GB