mtop

htop for Ollama — a curses-based TUI that monitors your models, GPU, and the Ollama server (in Docker or bare-metal) in real time. Zero flicker. Zero dependencies beyond Python 3.10+.

Why?

There are web dashboards, Prometheus exporters, and chat TUIs for Ollama. But there's no terminal monitor — something you SSH into a box and just run, like htop or nvtop, to see what models are loaded, how much VRAM they're eating, and whether the container is healthy.

mtop fills that gap. One file, one command, pure stdlib Python. It auto-detects whether Ollama runs in a Docker container or as a bare-metal process (systemd or a manual ollama serve) and monitors it either way.

Features

• Zero-flicker display — curses double-buffered rendering, no clear + print loops
• Effective inference config — RUNNERS section parses each runner's argv: context, batch, flash attention, KV cache dtype, layer offload. The only place these are observable
• Whole-tree process stats — the server and its per-model ollama runner subprocesses, with PSS accounting where available
• Loaded models — name, VRAM/RAM split, context length, processor type, TTL countdown
• Container health — status indicator (●/✗/○), uptime, CPU & memory with progress bars
• Multi-vendor GPU monitoring — NVIDIA via nvidia-smi and AMD via sysfs, side by side on the same host; utilization, VRAM/GTT, temperature, power draw
• AMD without ROCm — telemetry comes from /sys/class/drm/card*/device, so a bare amdgpu driver is enough; rocm-smi is only a fallback
• Jetson / Tegra / NVIDIA Spark — automatic fallback to unified memory via /proc/meminfo
• Non-blocking UI — all I/O (docker, nvidia-smi, HTTP) runs in a background collector thread; the interface stays responsive at 100 ms even when the API hangs, and stale data is flagged
• Interactive — q to quit, +/- to adjust refresh interval, o to toggle raw ollama ps
• Scriptable — --json one-shot mode for cron, Prometheus textfile collectors, or Ansible facts (exit code 1 on unhealthy)
• API-only mode — --no-docker for monitoring remote Ollama instances without local docker calls
• cgroup-aware CPU bar — normalizes against the container's --cpus/quota limit, not the host core count
• Docker and bare-metal — auto-detects the source: a Docker container, a systemd ollama.service, or a manual ollama serve; monitors process CPU/MEM via /proc (Linux, no root) or ps/sysctl (macOS)
• Docker-aware — talks to both the Ollama API and docker exec ollama ps
• Respects $OLLAMA_HOST — works with remote Ollama instances out of the box
• Zero external dependencies — only Python stdlib (curses, urllib, json, subprocess)

Quick Start

One-liner (no install)

curl -fsSL https://github.com/Quaerendir/mtop/releases/latest/download/mtop.py -o mtop.py
chmod +x mtop.py
./mtop.py

pip install

pip install ollama-mtop
mtop

From source

git clone https://github.com/Quaerendir/mtop.git
cd mtop
pip install -e .
mtop

Run directly from a clone (no install)

git clone https://github.com/Quaerendir/mtop.git
cd mtop
PYTHONPATH=src python -m mtop

Usage

mtop [-c CONTAINER] [-i INTERVAL] [-u URL] [-m MODE] [--no-gpu] [--no-docker] [--json] [-V] [-h]

Options:
  -c, --container NAME   Docker container name (default: ollama)
  -m, --mode MODE        Data source: auto|docker|local|api (default: auto)
                         local = bare-metal `ollama serve` (systemd/proc/ps)
                         api   = models only, no host resource stats
  -i, --interval SECS    Refresh interval in seconds (default: 1.0)
  -u, --api-url URL      Ollama API base URL (default: $OLLAMA_HOST or http://localhost:11434)
                         Scheme-less values (gpu-rig:11434) are accepted, like Ollama itself
      --no-gpu           Disable GPU monitoring section
      --no-runners       Hide the RUNNERS section (effective inference config)
      --no-docker        API-only mode: skip all docker calls (remote instances)
      --json             Print one snapshot as JSON and exit (exit 1 on unhealthy)
  -V, --version          Show version
  -h, --help             Show help

Examples

# Monitor a custom container name
mtop -c my-ollama

# Slower refresh for remote/metered connections
mtop -i 5

# Bare-metal Ollama (systemd service or `ollama serve` in a terminal)
mtop --mode local

# Let mtop figure it out (docker? systemd? manual? — it probes in that order)
mtop

# Monitor a remote Ollama instance — API only, no local docker/GPU noise
mtop -u 192.168.1.100:11434 --mode api

# One-shot health/state snapshot for scripting
mtop --json | jq '.models[].name'

# Using OLLAMA_HOST environment variable
export OLLAMA_HOST=http://gpu-rig:11434
mtop

Interactive Keys

Display Layout

─── mtop v0.2.0 — Ollama Model Monitor ───
host: gpu-rig     container: ● ollama     up: 3d 14h     2026-03-11 15:42:01
────────────────────────────────────────────────────────────────────────────────
CONTAINER RESOURCES
   CPU  [████░░░░░░░░░░░░░░░░░░░░░░░░░░]  12.3%
   MEM  [██████████████░░░░░░░░░░░░░░░░]  45.2%     14.2GiB / 31.4GiB
GPU
   [0] NVIDIA GeForce RTX 4090  42°C
     UTIL [████████░░░░░░░░░░░░░░░░░]  32.0%
     VRAM [██████████████████░░░░░░░]  72.4%     17382 / 24000 MiB
LOADED MODELS
   MODEL                                VRAM        RAM         CTX       PROCESSOR       EXPIRES
   ──────────────────────────────────────────────────────────────────────────────────────────────
   qwen2.5-coder:32b-instruct-q8_0     18.42 G     0.00 G      32768     GPU             4m 32s left
OLLAMA PS (raw)
   NAME                                SIZE       PROCESSOR    UNTIL
   qwen2.5-coder:32b-instruct-q8_0     19.8 GB    100% GPU     4 minutes from now

Supported Platforms

Requirements

• Python 3.10+ (uses match-era type hints like list[str], X | Y)
• Docker (for container monitoring)
• Ollama running in a Docker container (or accessible via API)
• nvidia-smi (optional, for GPU stats)

Roadmap

• Record terminal sessions with asciinema for README gif
• AMD GPU support (sysfs first, rocm-smi fallback)
• Apple Silicon GPU stats (via powermetrics)
• Model pull progress tracking
• Multiple container / multi-host support
• Configurable layout (raw ollama ps toggle; more sections to follow)
• Model actions — unload on keypress (keep_alive: 0), extend TTL
• Sparkline history for CPU/GPU utilization (braille chars, stdlib deque)
• systemd/bare-metal Ollama support (process stats via /proc, no Docker required)
• Log panel (tail Ollama container logs)
• Effective inference config per runner (context, flash attention, KV dtype)
• Request rate / tokens-per-second from Ollama API

Contributing

PRs welcome. Keep it stdlib-only — the zero-dependency constraint is a feature, not a limitation.

git clone https://github.com/Quaerendir/mtop.git
cd mtop
pip install -e .
# hack on src/mtop/*.py
mtop

# regenerate the single-file artifact shipped with releases
python tools/bundle.py       # -> dist/mtop.py

dist/mtop.py is generated — never edit it by hand.

License

MIT — see LICENSE.

Acknowledgements

Built as a collaboration between a human homelab geek and Claude (Anthropic) during a late-night infrastructure session. The original bash prototype migrated to Python/curses because fighting tput and jq in a loop was getting old.