Which local LLM can I run?
Start with RAM, VRAM, machine type, and workload, then test the top Fits rows.
Which local LLM can I run?
Find which local LLM your computer can run
Choose RAM, GPU memory, machine type, and workload. Click once to run the WhichLLM backend and get practical local LLM picks, fit counts, and install hints.
Ready to calculate
Run the finder
Choose a setup and click once. Backend results appear here.
Semantic tool entry
This is the direct Which Local LLM Can I Run page.
Use this page when the question starts with your computer: 8GB RAM laptop, MacBook, RTX 4060 Ti, RTX 4090, CPU-only box, or a model such as Qwen3, DeepSeek, Gemma, or Llama.
Which Ollama model should I install?
Use the calculator first, then copy the install command from a model that fits comfortably.
Should I upgrade my GPU for local AI?
Compare current hardware with RTX and workstation presets before spending money.
What hardware does one model need?
Use the target-model planner for Qwen3, DeepSeek, Gemma, Llama, and other local models.
All local model guides
Open the full local AI model hub after the calculator narrows the model range.
OpenChoose the local AI app
Decide whether to install Ollama, LM Studio, Jan, or Open WebUI first.
Open8GB RAM app path
Start safely on low-memory laptops before downloading a large model.
OpenFit method
Check how AI Jupyter separates fit, comfort, adoption, and real tests.
OpenContinue after the calculator
32B model
Qwen3 32B
A workstation-grade RAM and VRAM guide for Qwen3 32B.
32B model
DeepSeek-R1 Distill Qwen 32B
A 24GB VRAM local reasoning path for DeepSeek-R1 Distill Qwen 32B.
70B model
Llama 3.3 70B
A large-model planning guide for running Llama 3.3 70B locally.