verifier.org

llama.cpp

mit, and the engine most of this list is quietly built on top of.

free#anyone-comfortable-at-a-#anyone-who-wants-the-new

the actual software doing the work in ollama, lm studio, jan and koboldcpp — unbeatable on licence, backends and speed of support, and not a product for most people.

mit, no conditions, and the widest hardware support here: cuda, metal, rocm/hip, vulkan, sycl and cpu, on desktop, mobile and embedded builds. it is also first to support new model architectures and quantisation formats, usually within days, which is why every wrapper on this list depends on it.

it ships a cli, a server with an openai-compatible api, and a deliberately minimal built-in web ui. that is enough to run anything, and it is not an experience most people want as their daily driver.

the description above is ours, condensed from the ranking. pricing moves — check it on the vendor's own page before you rely on it.

pricing
free
website
github.com
our verdict

the actual software doing the work in ollama, lm studio, jan and koboldcpp — unbeatable on licence, backends and speed of support, and not a product for most people.

we researched this category against vendors' own pricing pages and licence files. that is where this line comes from — not from the vendor, and not from anything they paid for.

more local llm tools

Text Generation WebUI

oobabooga's gradio interface for running local models and loaders

#open-source#gradio

vLLM

high-throughput serving engine you can run on your own gpu

#serving#throughput

MLX LM

apple's framework for running and fine-tuning models on apple silicon

#apple-silicon

Jan

we checked thisfree

gui, server and inference engine all apache 2.0 — the only tool here where every layer is genuinely open.

#anyone-who-wants-ollama-