Runs open-weight models locally on kernels tuned to your hardware and connects them to coding agents through the magnitude CLI and an OpenAI- and Anthropic-compatible API.
Use it now
Keep it with up to 9 other tools you're setting up, then share them all as one link.
Try this first
Ask your coding agent:
- Hardware Model Recommendations
Use Magnitude to inspect local hardware: ‹what you need›
Install the CLI
Install it as its own instructions say.
Its page: magnitude.dev
Call the API
Auth None needed
Its page: magnitude.dev
Compare
Press Compare to add it to the tray at the bottom of the window, then pick up to 2 others anywhere on the site.
- Access
- CLIAPI
- Auth
- None
- Skills
- 4
Skills
Hardware Model Recommendations
Inspects local hardware and ranks compatible catalog models by preference with magnitude catalog recommendations, then pulls one to run.
Agent Connections
Configures Pi, OpenCode, Hermes, OpenClaw, Codex, Claude Code, Oh My Pi, or Cline to use local models with magnitude connections add.
Local Inference API
Serves OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages routes on port 10100, as JSON or server-sent event streams.
Remote Inference Server
Runs magnitude serve on a dedicated machine behind an API key so agents on another computer can use its models over the network.