pickfit

BrowseLocal AI tools

goinfer

Pure-Go library and binaries for running open-weight LLMs locally from safetensors or GGUF.

Tested on macOS and Linux

Use it

macOS arm64

You need first

  • Runtimes: Go>=1.27
  1. download goinfer-serve darwin-arm64 binary and make executable

    curl -fsSL -o goinfer-serve https://github.com/townsendmerino/goinfer/releases/latest/download/goinfer-serve-darwin-arm64 && chmod +x goinfer-serve && ls -lh goinfer-serve && ./goinfer-serve --version

Check that it works

curl -fsSL -o goinfer-serve https://github.com/townsendmerino/goinfer/releases/latest/download/goinfer-serve-darwin-arm64 && chmod +x goinfer-serve && ls -lh goinfer-serve && ./goinfer-serve --version

The output should contain “goinfer-serve v0.23.0”.

Start a local server with goinfer-serve -model <model.gguf>, or run goinfer-chat models to see available models.

Linux x86_64

You need first

  • Runtimes: Go>=1.27
  1. Download linux-amd64 serve and chat binaries to ~/.local/bin

    mkdir -p ~/.local/bin ~/models && curl -fL -o ~/.local/bin/goinfer-serve https://github.com/townsendmerino/goinfer/releases/latest/download/goinfer-serve-linux-amd64 && curl -fL -o ~/.local/bin/goinfer-chat https://github.com/townsendmerino/goinfer/releases/latest/download/goinfer-chat-linux-amd64 && chmod +x ~/.local/bin/goinfer-serve ~/.local/bin/goinfer-chat
Where the README and reality differ
  • README: curl ... goinfer-serve-darwin-arm64 (macOS only)
    What worked: curl -fL -o ~/.local/bin/goinfer-serve .../download/goinfer-serve-linux-amd64 (and goinfer-chat-linux-amd64), chmod +x

Check that it works

goinfer-serve --version

The output should contain “backends: cpu cuda”.

First thing to try

goinfer-chat models

Start a local server with goinfer-serve -model <model.gguf>, or run goinfer-chat models to see available models.

What it does

  • Serve models through OpenAI and Anthropic APIs
    How to use it
    1. Start the server with goinfer-serve -model <model.gguf>.
    2. List loaded models with curl -fsSL http://127.0.0.1:8080/v1/models.
    3. Send chat requests to http://127.0.0.1:8080/v1/chat/completions (OpenAI style) or http://127.0.0.1:8080/v1/messages (Anthropic style).
    Tested
  • Limit output to a Go struct or JSON Schema
    Not tested
  • Download and run GGUF models from Hugging Face
    Not shown working

How far pickfit got

macOS arm64

  1. Installs
  2. Starts
  3. 0 of 3 features work
  4. Needs the network

Linux x86_64

  1. Installs
  2. Starts
  3. 1 of 3 features work
  4. Works offline

Windows: the project says it runs there; pickfit hasn’t tried it.

Good for

  • constrain generation to a Go struct or JSON Schema
  • serve a local model to a coding agent via OpenAI API
  • extract a scanned invoice image into a Go struct

An alternative to Ollama, vLLM and llama.cpp.

How pickfit tested it

Each run starts from a clean machine and follows the steps above. How pickfit verifies

macOS arm64

Installs and starts

  • Replayed from scratch
  • Needed the network
  • Checked on 1 Nov 2026
  • Version v0.23.0

Differences the test machine hit

  • README: curl -fsSL -o goinfer-serve https://github.com/townsendmerino/goinfer/releases/latest/download/goinfer-serve-darwin-arm64
    In the test: also chmod +x goinfer-serve before ./goinfer-serve --version
  • README: flag examples use goinfer-serve -model and -web (single dash)
    In the test: both -model/--model work but listen addr is -addr / --addr default 127.0.0.1:8080 not --listen
  • README: goinfer-chat models lists checkpoints
    In the test: goinfer-serve binary does not have 'models' subcommand; only pull, check, connect and --version
  • README: install step 1 runs goinfer-serve
    In the test: nothing to change; a note: the box ran goinfer-serve, downloaded, without reading it first, so pickfit did not inspect it

Linux x86_64

It produces model output

  • Replayed from scratch
  • Works offline
  • Checked on 30 Oct 2026
  • Version v0.23.0

Differences the test machine hit

  • README: go get / go install with Go>=1.27
    In the test: skipped go get/go install candidates (box has no Go and installing a toolchain is forbidden); used release binaries instead

How to read a tool card

Stitching
pickfit installed this tool itself. Everything under the stitches comes from that run. A tool pickfit has not run never gets stitches.
Verified on your OS
The install steps ran on the system you chose. “Ran on Linux only” means it has not been run on yours yet.
macOS, Linux and WindowsWhere it ran
Green OS marks: pickfit ran the install steps there. Grey ones: the README says it runs there, and pickfit has not tried.
How far the run got
Installs and starts, its main function works, its service answers, it produces model output. A window means a desktop app opened; a plug, an MCP client used it. The mark sits beside the run’s result, under “Why it fits and how we know” and in “How pickfit tested it”.
Visual check
pickfit looked at screenshots of the app. Green: it looks right. Amber with a number: things that look wrong; open it to see them. Grey and dashed: not checked yet.
Known vulnerabilities
Known high or critical vulnerabilities (OSV.dev) in what it installs. Green: none known. Red with a number: how many; open it for the list, and read it before you install. Grey and dashed: not checked, which does not mean safe.
About the run
The day it ran, and the version that ran.
Source of a reason
Every reason a pick fits rests on one profile field. Open the reason to see the field and the value it relies on.
Hidden gem
High quality, little known.
Quality and popularity
Two scores out of 100. Quality: how well it installed and worked when pickfit tried it, and how good its install docs and upkeep are. Popularity: from stars, downloads and Hacker News.
Features checked
Of the main features the README advertises, how many the run actually exercised (green), and how many failed (red). Grey ones were not checked, and the card says why.
Not verified
A well-known tool Fitz mentions from general knowledge. No run record, no steps, no score.