Open-source AI models
Open-source models are free to download and run on your own computer — no API key, no monthly fee, and nothing sent to anyone else's server. The catch is memory: a model has to fit in your RAM. Of the 28 language models we track, 19 run on ordinary consumer hardware, plus 2 image generators.
What can your computer run?
Two questions: how much memory you have, and what kind. The first decides what loads, the second decides how fast it runs.
1. How much memory?
2. What kind of machine?
On a Desktop or laptop — DDR5-5600 with 16GB RAM, 8 of 28 models fit. The most popular:
- Qwen3.5-9B9.7B6.8 GB~10 tok/sComfortable
- gpt-oss-20b22B13 GB~25 tok/sFast
- Qwen3.5-4B4.7B3.8 GB~21 tok/sComfortable
- Qwen3.5-0.8B873M1.3 GB~110 tok/sFast
- Qwen3.5-2B2.3B2.0 GB~46 tok/sFast
Memory figures are Q4_K_M at 4K context, including the context cache and runtime overhead. Speed is calculated, not measured: 90 GB/s × 0.65 ÷ bytes read per token.
Bandwidth source: 5,600 MT/s × 128-bit dual channel ÷ 8 = 89.6 GB/s. How these are worked out →
Browse by what it takes to run
Sorted by hardware, not by benchmark score.
Most-downloaded right now
What people are actually running this month, straight from HuggingFace.