Show HN: LocalCoder – Tell it your hardware, get the exact local AI model to run
Hey HN — I built this after seeing the Qwen3-Coder threads here. Every thread had the same questions: which quant for my GPU? How much VRAM do I need? Ollama or llama.cpp? What context window can I actually use?
LocalCoder answers all of that in one page. Pick your platform (Apple Silicon, NVIDIA, CPU), select your chip and memory, and it gives you:
- The best model + quantization for your setup - Expected speed (tokens/sec) and context window - Copy-paste Ollama commands to get running in 60 seconds
The recommendation engine is a curated config matrix built from HN benchmarks, Unsloth docs, and llama.cpp test data. No AI inference on the backend — it's all client-side.
Free tier gives you the top pick + Ollama commands. $9 one-time unlocks alternatives table, llama.cpp commands, and IDE integration guide.
Would love feedback on the recommendations. If your hardware isn't covered or a rec seems off, let me know — I'll update the matrix.
We took this a step further with ClawBox (https://openclawhardware.dev): instead of telling users which model to download, we ship them hardware with the right models already installed and optimized. It is a Jetson Orin Nano 8GB + 128GB NVMe with OpenClaw pre-configured — plug it in, connect to your network, and you have a local AI assistant running in minutes.
The challenge with recommendation engines is that even when you tell someone "use Q4_K_M Mistral 7B on your RTX 3060," they still need to install ollama, configure VRAM allocation, deal with context length settings, etc. For non-technical users, the gap between "know which model" and "actually running it" is still huge.
That said, your tool is really valuable for the tinkerer crowd who enjoys the setup process. Different audiences, complementary approaches.