llmfit

Summary

Hundreds of models & providers. One command to find what runs on your hardware.

Command line

Values are placeholders taken from each option’s declared default. Nothing is executed here — the output shown is a recording of a run that already happened.

Release history

1. v1.1.10 Aug 17, 2026 · issue 007

   llmfit v1.1.10 adds RamaLama runtime discovery via MCP and expands the model catalog with the Qwen3.8 family.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch v1.1.10 https://github.com/AlexsJones/llmfit.git
   # already have the repo? check out this version:
   $ git checkout v1.1.10

   └──▷ TRY IT
   See how the new Qwen3.8 models rank against your hardware, filtered to good-or-better fits.
   copy

   $ llmfit recommend --json --limit 5 | jq '.models[] | select(.name | test("Qwen3"))'

   • ›Adds the Qwen3.8 model family to the built-in catalog, making Qwen3.8 variants scoreable and recommendable against local hardware.
2. v1.1.8 Aug 4, 2026 · issue 007

   llmfit v1.1.8 adds per-preset benchmark counts to the TUI hardware picker.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch v1.1.8 https://github.com/AlexsJones/llmfit.git
   # already have the repo? check out this version:
   $ git checkout v1.1.8

   • ›Displays per-preset benchmark counts in the TUI hardware picker, so you can see at a glance how many community measurements back each hardware preset.
3. v1.1.7 Aug 3, 2026 · issue 007

   llmfit v1.1.7 adds a --providers filter to fit, RamaLama store discovery, and GGUF-source row markers in the TUI.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch v1.1.7 https://github.com/AlexsJones/llmfit.git
   # already have the repo? check out this version:
   $ git checkout v1.1.7

   └──▷ TRY IT
   Filter fit results to models from specific GGUF providers to find what unsloth or bartowski quantizations will run on your hardware.
   copy

   $ llmfit fit --providers unsloth,bartowski -n 10

   Pipe provider-filtered fit results as JSON into jq for scripted or agent consumption.
   copy

   $ llmfit fit --json --providers unsloth,bartowski --perfect

   • ›Adds --providers flag to llmfit fit, accepting a comma-separated, case-insensitive list of provider names (e.g. unsloth,bartowski) to narrow fit results to specific canonical or GGUF providers.
   • ›Detects RamaLama store models via ramalama ls when no server is running, so locally stored models appear in fit results without requiring an active runtime.
   • ›TUI now marks list rows matched through a GGUF source, making it visually clear which results come from GGUF-backed providers.
   • ›TUI Fit box now spells out active advanced range filters (e.g. Too Tight ≤2B, mem≥50%) in warning color instead of showing cryptic R/M markers, and the empty-results hint names the active filter explicitly.