LM Studio
Desktop GUI for running local LLMs.
FreemiumCPU-capable
Actually FreeNo SignupWatermark-FreeHobbyist-Friendly
Jan
Open-source ChatGPT desktop — runs models locally or via API.
Open SourceCPU-capableRuns locally
Actually FreeNo SignupOpen SourceWatermark-FreeHobbyist-FriendlyAPI
Visit JanUpdated 2026-05-12 · Direct link
Hardware requirements
Runs locally · Entry GPU (6–8 GB)
Min VRAM
None
Rec. VRAM
8 GB
Min RAM
8 GB
Rec. RAM
16 GB
Disk
20 GB
GPU class
Entry GPU
Apple Silicon ✓CPU-CapableQuant: Q4_K_M, Q5_K_M, Q8_0 +1
CPU-only viable for 7B models; GPU strongly recommended for 13B+.
Screenshot placeholder · Jan
What is Jan?
Jan is the closest open-source equivalent to a polished ChatGPT desktop client. Built on llama.cpp under the hood, it ships with a model hub UI, multi-model conversation tabs, an OpenAI-compatible local API server, and remote-model support (OpenAI, Anthropic, Groq) gated behind your own keys. Cross-platform and AGPL-licensed.
Pros & cons
Pros
- Native Electron app for Win / Mac / Linux — no terminal
- Local OpenAI-compatible server out of the box
- Built-in model hub with one-click downloads
- Bring-your-own-key support for remote models
Cons
- Electron means ~200 MB RAM overhead before any model loads
- Plugin / extension API is still maturing
What's actually free?
AGPL desktop app; free for personal & commercial use.
✓ Actually FreeNo SignupOpen SourceWatermark-Free
Alternatives
Ollama
One-command local LLM runtime.
Open SourceCPU-capable
Actually FreeNo SignupOpen SourceWatermark-Free
Msty
Polished local-LLM client with split chats and knowledge stacks.
Freemium · from $4.16/moCPU-capable
Actually FreeNo SignupWatermark-FreeHobbyist-Friendly
AnythingLLM
RAG-first local LLM workspace with workspaces and agents.
Open SourceCPU-capable
Actually FreeNo SignupOpen SourceWatermark-Free