Reviewed category
Local LLM Tools
Choose a desktop app, model runner, or self-hosted interface for using language models on your own hardware.
- 6
- reviewed services
- Manual
- publication review
- 2026-10-03
- latest review date
Use with context
What to know before choosing
A local LLM runs a language model on a computer you control. The model is the downloaded set of weights; the tools below help load it, chat with it, or connect it to another application. They are not interchangeable subscriptions to the same AI.
Choose your starting point
- A desktop chat app: LM Studio, Jan, or GPT4All. Start here when you want to try a model without building a chat interface.
- A model runner: Ollama for managing a local model workflow, or llama.cpp for more hands-on runtime configuration.
- A browser interface for a model backend: Open WebUI. It still needs a configured model connection; the interface alone is not a language model.
Before downloading
Check available memory, storage, operating-system support, and the selected model’s license. A model that downloads successfully may still be too demanding for comfortable use. Try a small, non-sensitive task before committing to a larger setup; hardware, electricity, hosting, and optional cloud services can all add costs.
Local-capable does not mean every feature stays local. Check which model endpoint is selected and whether cloud models, web search, connectors, telemetry, or sharing options are enabled. Keep a local server private unless you have configured appropriate access controls. Treat third-party extensions and model downloads as software you need to trust.
Review generated answers and code, protect private documents and secrets, and follow workplace policy. These pages are official-source reviews, not hardware benchmarks or security certifications. Browse all reviewed tools for hosted alternatives.
Fair comparison
Try every candidate on the same task
Try the same non-sensitive prompt on the same model and computer before comparing setups.
Check consistently: Memory use, response time, answer quality, model licenses, and local versus cloud connections.
Manual review index
Reviewed Local LLM Tools
Each card highlights one reviewed task-fit example as a starting point, not a ranking.
GPT4All
A desktop application for running local language models and exploring questions over selected local documents.
Example task Trying local chat in a desktop app ReviewedJan
A desktop AI app for local model chat, with optional connected providers that should be checked separately.
Example task Exploring a desktop local-chat workflow Reviewedllama.cpp
A hands-on inference runtime for running supported language models on local hardware or a server you manage.
Example task Learning how local model inference is configured ReviewedLM Studio
Use a desktop interface to download, manage, and chat with local language models; check separately for connected or cloud features.
Example task Trying local models in a desktop chat interface ReviewedOllama
Download and run language models locally, then connect them to a chat or development workflow. Cloud options are separate.
Example task Running a downloaded language model ReviewedOpen WebUI
A self-hosted chat interface for local or remote model backends. Configure the connection and data flow before uploading documents.
Example task Adding a browser interface to a model backend