omm setting
View or change telemetry, what anonymous data may leave this machine, memory protection, the update channel, the color theme, calibration, and signed-catalog trust — interactively, or one subcommand at a time.
01 / 07
Overview
Reach for setting any time you want to see or change how omm behaves outside of a single command. Bare omm setting opens an interactive menu covering the same ground as every subcommand below; each subcommand also works directly, prints a small table of its current value either way, and every write here stays local — nothing is uploaded just by changing a setting. Everything omm may ever send is now grouped under one subcommand, omm setting upload, with a separate policy per channel: benchmark results, anonymous usage stats, and crash reports. Usage stats and crash reports are off until you turn them on.
02 / 07
Options
Every flag this command accepts, and what it defaults to when you leave it out.
setting telemetry --endpoint URL—Default: not configuredConfigure where benchmark telemetry is sent, or clear it with --endpoint none.
setting upload—Default: show all three policiesShow all three outbound-data policies — benchmark, usage, crash — in one table.
setting upload benchmark --enable|--disable|--ask—Default: askConfigure whether benchmark results upload without asking, never, or ask each time.
setting upload usage --enable|--disable|--reset-id—Default: offAnonymous daily usage stats: omm version, install method, OS, CPU/GPU class, RAM and VRAM size ranges, and which commands ran. Never model names, search terms, file paths, IP, or hostname. With no flag it prints the exact payload that would be sent next; --reset-id generates a new random install id.
setting upload crash --enable|--disable|--ask—Default: offConfigure the separate, opt-in policy for scrubbed crash reports, sent on their own write-only channel.
setting memory-guard --policy ask|block|observe --poll-seconds N --low-memory-seconds N—Default: show current valuesConfigure Memory Guard's policy, polling interval (0.1–60 seconds), and how long low memory must persist before cancellation (0–300 seconds).
setting version --stable|--beta—Default: stableShow or switch which branch omm update pulls from.
setting theme --set NAME—Default: darkShow or change omm's output color theme.
setting calibrate [name]—Default: smallest Ollama modelCorrect this machine's local speed prediction using a real Ollama-linked model — never uploaded.
setting catalog-trust --manifest-url URL --public-key KEY—Default: unsetRequire future recommendation downloads to pass Ed25519 signature verification.
setting catalog-status—Default: —Show the current signed-catalog trust and rollback state.
setting catalog-rollback—Default: —Restore the most recent different recommendation snapshot.
03 / 07
Examples
From a plain search to something you'd put in a script.
Open the interactive menu covering every setting below.
$ omm settingShow the current color theme.
$ omm setting themeShow all three outbound-data policies at once.
$ omm setting uploadTurn anonymous usage stats off and discard anything already queued.
$ omm setting upload usage --disableShow signed-catalog trust and rollback state.
$ omm setting catalog-status04 / 07
A real run
Real bare omm setting upload run, 2026-09-01, against a throwaway OMM_HOME — the three real default policies on a machine that has never been asked, verbatim including the hint line under the table. Table borders are dropped for the web, as everywhere on this page set.
06 / 07
If something goes wrong
Every message below is one this command actually prints. Find yours, read why it happened, then do the last line.
Choose only one of --enable, --disable, or --ask.- why
- omm setting upload benchmark and omm setting upload crash each take exactly one policy flag; this run passed more than one at once.
- what to do
- Pass only one of --enable, --disable, or --ask.
- source
- src/omm/cli.py:6454, :6489
Set an endpoint with `omm setting telemetry --endpoint` before enabling uploads.- why
- Benchmark uploads have nowhere to go until a telemetry endpoint is configured, so --enable refuses rather than silently enabling a policy that can never send.
- what to do
- Run omm setting telemetry --endpoint URL first, then re-run the --enable.
- source
- src/omm/cli.py:6459-6460
The signed catalog manifest must use HTTPS.- why
- Signed-catalog verification has to happen over a connection that can't be tampered with in transit — an HTTP manifest URL defeats the point of requiring a signature at all.
- what to do
- Use an HTTPS manifest URL.
- source
- src/omm/cli.py:6828
--policy must be ask, block, or observe.- why
- Memory Guard's policy only has three real values, and this run passed something else.
- what to do
- Use ask, block, or observe.
- source
- src/omm/cli.py:6602
Still stuck? Open an issue with the exact message you saw.
07 / 07
CLI reference
Exactly what omm setting --help prints, exported from the CLI source.
Usage
omm setting [OPTIONS] COMMAND [ARGS]...Options
This command takes no options of its own.
Sub-commands · 11 sub-commands
omm setting auto-import
Automatically adopt models that Ollama, LM Studio, and similar apps download natively into the omm hub in the background. Off by default. See PRIVACY.md.
omm setting auto-import [OPTIONS] COMMAND [ARGS]...This command takes no options of its own.
omm setting calibrate
Correct this machine's local speed estimate without uploading data.
omm setting calibrate [OPTIONS] [model_name]This command takes no options of its own.
omm setting catalog-rollback
Restore the most recent different recommendation snapshot.
omm setting catalog-rollback [OPTIONS]This command takes no options of its own.
omm setting catalog-status
Show recommendation-catalog trust and rollback state.
omm setting catalog-status [OPTIONS]This command takes no options of its own.
omm setting catalog-trust
Require future recommendation downloads to pass signature verification.
omm setting catalog-trust [OPTIONS]--manifest-url <str>Default: —HTTPS manifest URL.
--public-key <str>Default: —Base64 Ed25519 public key.
omm setting memory-guard
Show or change the consent-aware runtime memory protection policy.
omm setting memory-guard [OPTIONS]--policy <str>Default: —Memory Guard policy: ask, block, or observe.
--poll-seconds <float range>Default: —Seconds between live-memory checks during a long operation.
--low-memory-seconds <float range>Default: —How long low memory must persist before OMM cancels its own operation.
omm setting runtime-profile
Inspect a saved runtime profile, or undo the last save without reloading models.
omm setting runtime-profile [OPTIONS] {model_name}--engine <str>Default: ollamaollama or lmstudio
--restoreDefault: —Restore the previous saved profile (or defaults).
omm setting telemetry
Configure where benchmark telemetry is sent; see `omm setting upload` for the send policy.
omm setting telemetry [OPTIONS]--endpoint <str>Default: —Self-hosted HTTPS endpoint, localhost URL, or 'none' to clear it.
omm setting theme
Show or change the color theme applied to omm's output.
omm setting theme [OPTIONS]--set <str>Default: —One of: light, dark, high-contrast, no-color
omm setting upload
Choose what anonymous data omm may send: benchmark results, usage stats, crash reports. Each is off or ask by default. See PRIVACY.md.
omm setting upload [OPTIONS] COMMAND [ARGS]...This command takes no options of its own.
omm setting version
Show or switch the update channel `omm update` pulls from. Switching takes effect immediately - it fetches and checks out the new branch right away, no separate `omm update` needed.
omm setting version [OPTIONS]--stableDefault: —Track the stable channel (main branch).
--betaDefault: —Track the beta channel (beta branch).
Shared flags
Every omm command also accepts --json, --no-color, --quiet, -q, --yes, -y, so they are listed here once instead of on each command.
Exported from omm 0.3.101.
All commands
- omm searchFind a model across the curated catalog, HuggingFace and ModelScope.
- omm installDownload a model into the hub and link it into every installed runner.
- omm runChat with an installed model — in the terminal for Ollama, or by opening the app for GUI runners.
- omm recommendGet a model suggestion ranked for this machine's hardware, with an offer to install it.
- omm contributeBenchmark models in a loop, uploading telemetry to improve recommend for hardware like yours.
- omm setupRe-run the hardware scan and runner-install checklist, any time.
- omm scanPrint this machine's hardware, detected runners, and models — no flags needed.
- omm tuneGet recommended context length, GPU offload, threads, and batch size for a model.
- omm fitSee whether a model fits this machine's free memory right now, installed or not.
- omm helpShow omm's own command summary, or the full reference with --all.
- omm importAdopt .gguf files sitting in other apps' model directories into the omm hub.
- omm uninstallRemove a model and clean up its symlinks and manifests. Alias: rm.
- omm listShow every model omm has installed and which runners each is linked into. Alias: ls.
- omm infoShow full detail — repo, version, size, links, run commands — for one installed model.
- omm upgradeRefresh installed models against their source — only re-downloads what's actually changed. Alias: up.
- omm linkRe-verify and repair every installed model's runner links, or link into a custom directory.
- omm cleanupClean up leftover partial downloads and broken runner symlinks in one pass — no flags needed.
- omm verifyProve that an installed model actually loads and generates text on this machine.
- omm benchmarkLocal quality and speed smoke evidence for one or more installed models.
- omm updateReinstall omm from the latest source and refresh its recommendation data.
- omm doctorDiagnose the omm install and Ollama links, read-only — no flags needed.
- omm engine installInstall one local AI runner program directly, skipping the setup checklist.
- omm logRead the local run log: what omm ran, when, and whether it worked.
- README — UsageEvery omm command, one line each.