Skip to Content
Walkthrough

5-minute walkthrough

From a fresh install to your first alert. You can stop at step 2 if all you want is the free health check.

1. Health check

From any repo:

mm status

The CLI pulls a signed registry snapshot from cdn.llmstatus.ai, scans the directory on-device, and prints every AI model it finds along with its lifecycle status and the soonest retirement date.

LLM Status — registry 20260528T060033Z (today), 380 models
Scanned ./apps/web: 12 reference(s) → 6 model(s), 1 custom

Models in use:
  🔴 retired     openai/gpt-4-0314          retires 2024-06-13   → openai/gpt-4-1  (2)
  🟠 retiring    anthropic/claude-opus-4    retires 2026-06-15   → anthropic/claude-opus-4-7  (1)
  🟢 ok          openai/gpt-5                                                       (3)

⚠ 2 model(s) need attention before they retire.

No telemetry, no account, fully offline-capable after the first run. The legend:

  • 🟢 ok — current model, no action needed
  • 🟡 deprecating — provider has stopped recommending it, no firm retirement date yet
  • 🟠 retiring — firm retirement date within the next 90 days
  • 🔴 retired — retirement date has passed; calls may already be failing
  • ⚪ custom — string didn’t match anything in the registry (custom or fine-tuned model)

2. Sign in

If you want alerts on deprecations / retirements across your team, or an inventory of models across repos, sign in. Opens a browser tab — no terminal-paste of API keys needed:

mm login

The CLI generates a one-time device code, opens llmstatus.ai/cli?code=... in your browser, you confirm in your account, and the key gets written to ~/.config/llmstatus/config.json (mode 0600). To paste an existing key instead:

mm login mm_live_yourKeyHere

3. Scan a project

One scan finds every model call, lets you pick what to track, and uploads to your account:

mm scan                              # interactive TUI
mm scan --ci --json --project web    # for pipelines
mm scan --sources all                # include env, k8s, helm, sql, aws-secrets

Or bare mm launches the full TUI (inventory · scan · what’s new · alerts) if you’d rather browse.

The detection sources:

SourceWhat it reads
filesystemrepo files (default)
envlive process env vars (OPENAI_API_KEY, etc.)
aws-secretsAWS Secrets Manager + SSM
k8skubectl secrets + configmaps
helmhelm release values
sqlpsql, via --db DSN

Privacy: the secret-aware sources shell out to your already-authenticated CLIs, run read-only, redact every snippet, and only ever upload model ids. Secret values never leave your machine. Use --dry-run to preview what would upload before it does.

4. Set up alerts

In the web app: Notifications → Add rule. Pick:

  • Scope — my models (only what you’ve added), providers (e.g. all OpenAI), or all (the whole registry).
  • Event types — deprecation, retirement, new model, replacement available.
  • Severity — info, warning, urgent.
  • Lead-time chips — 90 / 30 / 7 / 1 day before retirement.
  • Channels — email, Slack, SMS, in-app.
  • Delivery — immediate or daily digest.

Notifications are deduped per (user, dedupe_key) so one alert covers all your usages of a model — no spam.

Upgrade from the CLI: mm upgrade opens Stripe checkout and polls until Pro is active.

5. Drop into CI

Catch regressions when someone pins a deprecated model:

# .github/workflows/mm.yml
- run: npx @modelstatus/cli scan --check

Fails the build (non-zero exit) if any usage maps to a retiring or retired model. The CLI binary is also available on the CDN  — drop it into your Dockerfile if you’d rather not pull from npm in CI.

InstallCommands