Documentation · Reference

Free vs Pro comparison

Complete, factual breakdown of every capability included in STALMAS Free, unlocked with STALMAS Pro, and device hardware requirements.

Hardware notice: Available features depend on your device’s hardware. Requires Android 12+, a 64-bit CPU and 6/8 GB RAM for local AI, and Google Play Services (GMS) for Google Drive backups. See the full requirements table at the bottom of this page.

STALMAS Free covers complete day-to-day administration of up to 3 servers in total (Portainer and SSH, split however you like) with zero ads and no time limit. Pro unlocks unlimited servers, on-device AI, one-tap diagnostics, background alerts, extended terminal tools, and encrypted backups.

Price: Pro is a one-time payment of $14.99 USD base on Google Play, with no subscription and no renewal. Price varies by country and Google Play displays the final amount in your local currency. Pro users can also leave an optional tip ($2.99, $4.99, or $9.99 USD) from About (unlocks no extra features).

Connections

Feature Free Pro
Portainer and Linux SSH servers (password or .pem/.key private key) 3 in total, split however you like Unlimited
UP/DOWN status for every server Yes Yes
Allow self-signed TLS certificate (per Portainer server) Yes Yes

Portainer: dashboard, containers, and stacks

Feature Free Pro
Card dashboard (fixed): Health, Live resources (CPU/RAM/images), Top usage, Storage (system/df, reclaimable space), and Quick actions Yes Yes
Customizable dashboard: edit (reorder, hide, add cards), Health/Operations presets, and “Containers by stack” card No Yes
Pro dashboard cards: “All servers” (health across every saved server) and “AI diagnosis” No Yes
Pro cards: “Last 24 h” (host CPU/RAM with peak), “Activity” (background monitor findings even while the app is closed), “Trend” (container approaching its memory limit fastest), and “Activity” preset No Yes
Usage history: store CPU/RAM for up to 30 containers per environment on each check (7 days, encrypted, opt-in in Monitoring) No Yes
Container detail history: 6 h / 24 h / 7 d CPU and RAM charts with min, avg, and max (tracks the container even if recreated) No Yes
Start, stop, restart, pause, and resume containers Yes Yes
Container detail: color status header, link to its stack, Restart · Stop/Start · Console · Processes actions, and Summary · Logs · Usage · Config tabs Yes Yes
Summary: failure reason (healthcheck output, Out of memory with limit, explained exit code) with Restart / View errors; copyable IP, ports, restart count & policy, uptime, and CPU/RAM Yes Yes
Summary: “Notify me if it fails” per-container watch (alerts even when general monitoring is off; managed in Monitoring) No Yes
Summary: “Add to home screen” (preconfigured widget for that container) No Yes
Config: General, Resources, Healthcheck, Network & ports (tappable network → Networks), Volumes, Environment (secrets masked), and Labels Yes Yes
Rename container Yes Yes
Update container to latest image: pulls, compares, and recreates with the same configuration; for Portainer stack containers, offers redeploying the entire stack with fresh images Yes Yes
Kill and remove container confirming by name Yes Yes
Live usage: CPU/RAM, network, disk, PIDs charts, and running process list Yes Yes
Multi-select bulk actions: start, restart, or stop multiple containers Yes Yes
Containers: tappable filter counts (All / Healthy / Issues / Stopped), search by name, image, or port, Priority (“Needs attention”) and By-stack views, per-row CPU/RAM Yes Yes
Containers: “Explain” button on each issue (opens AI diagnosis with its logs) No Yes
Background alerts: notifies when a container stops, restarts unexpectedly, turns unhealthy, enters a restart loop, or an environment goes down (checks every 15–60 min, not real-time) No Yes
Home-screen widget: Summary (running / stopped / issues; active server) Yes Yes
Home-screen widget: single container with Restart and Refresh buttons, last-updated timestamp, and any saved server No Yes
Logs: ANSI colors, search, highlighted errors & warnings; live stream, filters (20 to 10,000 lines, preset or date range, timestamps), All / Errors / Warnings counts, copy, export .log / share, and “jump to first error” Yes Yes
In-container console (exec) Yes Yes
One-tap AI log analysis: sends only the currently filtered lines (max 100) and tells you what was sent; “Explain” filters Errors and analyzes No (paste manually in Free) Yes
Stacks: view compose & env, real container-level health bar, tappable filters, redeploy, start, stop, and delete (with confirmation), logs, and “Standalone containers” Yes Yes
Stacks: “Why is X failing?” button (AI diagnosis of the failing container) No Yes
Stack detail: Summary, Compose, Environment; Git (pull changes and configuration) Yes Yes
Create new stacks: Compose Editor, from Git (Source on Portainer 2.45+ or URL/token on older versions), and from Templates (Portainer & custom) Yes Yes
Analyze Compose file with AI and apply improvements straight into the editor No Yes
Storage: images and volumes (inspect, see what uses them, delete, reclaim space, pull latest image) Yes Yes
Storage: “What is safe to delete?” AI check No Yes
Resources tab: disk usage card, Clean up (prune images, volumes, networks, and build cache), and shortcuts to Images, Volumes, and Networks Yes Yes
Networks (read-only): search by network or container, All / Unused filter, copyable subnet, gateway, scope, and member containers Yes Yes

Terminal (SSH & container console)

Feature Free Pro
xterm-256color emulator with Esc, Tab, Ctrl, and arrow keys Yes Yes
Copy last command output Yes Yes
Per-server command history (stores up to 500) Views latest 50 Views all 500
Saved snippets Up to 10, no variables Unlimited, with {{name}} variables
Customizable key row with touch cursor pad No Yes
Keep screen awake Yes Yes
Terminal screen context for AI Up to 4,000 characters Up to 16,000 characters (cloud) · 1,500 (local)

AI Assistant

Shared core (included in Free and Pro)

Feature Detail / limit
Floating STALMAS assistant (character) Draggable bubble above keyboard, non-modal resizable anchored panel, in-panel history, and reply preview with ▶ / Apply
Cloud Gemini chat (BYOK) Uses your own API key, streaming with stop button and retry status; quota governed by your key
Gemini model selector Free/paid badges, dynamic list from your key, estimated tier checker, and 7-day local usage count
Privacy Gatekeeper (“redact and send”) Detects IPv4/IPv6 addresses, domains, passwords, keys, and connection strings; masks them with [REDACTED], sends, and shows a count with “view what was redacted”
Chat utilities Copy reply & code blocks, regenerate, edit last prompt, retry after error, CLOUD:/LOCAL: engine tag, temporary chat (not saved), search by title, and configurable retention
Per-server chat isolation Each chat stays tied to its server or environment and never mixes with others
Dangerous command confirmation Mandatory confirmation for destructive commands (rm -rf, docker system prune…), even with auto-run
Recent memory & manual context Up to 6 messages (3 exchanges) in Free; paste stacks or logs manually into chat
Report an AI response Flag button on every reply with automatic secret redaction to report@stalmas.app

Pro-only AI features

Feature Detail / limit
On-device local AI (LiteRT-LM in isolated :ai process) Download models (verified Gemma 4 E2B/E4B catalog, Hugging Face search, and local SAF import), preload, and Auto (GPU/CPU) or manual accelerator; 4,000-character shared window (maxNumTokens <= 4,000)
Auto mode & “Local AI first” Automatically routes prompts containing private data to your local model when installed
Reversible cloud pseudonyms (PseudonymMap) Gemini sees [IP_1] / [HOST_1] while you see your real IPs and hosts locally; secrets are never restored
Per-server notes & “Local AI only” rule Up to 1,500 characters of notes per server or environment
Custom skills (SKILL.md) Specialized instruction modules activated by context
Automatic stack & log reading (“Sees this screen” chip) and one-tap analysis Stacks: max 16,000 chars (head); logs: last 100 filtered lines & max 16,000 chars (tail) in cloud; local max 1,500 chars; truncation notice in chat
Extended cloud conversation memory Up to 20 messages (10 exchanges) in cloud; up to 6 messages in local within 40% of the local budget
Full chat history Full-text search inside messages, filter by server, pin, rename, and export to Markdown
Encrypted AI backup Chats, notes, and per-server policies included inside Pro encrypted backups

Settings, security, and data

Feature Free Pro
3-sheet welcome onboarding on first launch (skippable and replayable from About) Yes Yes
Spanish & English interface Yes Yes
Nord theme (light/dark) Yes Yes
OLED, Catppuccin, and STALMAS brand themes No Yes
PIN & biometric app lock (plus always-on SQLCipher + Keystore local encryption) Yes Yes
Encrypted backup & sync (AES-256) to Google Drive and local file No Yes
App Logcat viewer Yes Yes
Report a problem / technical error diagnostics (reviewable draft to report@stalmas.app) Yes Yes
About, open-source licenses, legal notices, and support Yes Yes
Support development (optional tip: $2.99 / $4.99 / $9.99 USD; unlocks no features) No Yes

Device requirements

Feature Device requirement What the app does if missing
STALMAS app Android 12+ (API 31 or higher) Google Play restricts installation to supported devices.
Local AI (engine & models) 64-bit CPU (arm64-v8a or x86_64), 6 GB RAM min (Gemma 4 E2B) or 8 GB RAM (Gemma 4 E4B), and free storage (model size + 500 MB) Without 64-bit, disables catalog, search, import, and activation with an explanation; if RAM or space is insufficient, blocks download with the reason or falls back to the previous valid model.
Biometric app lock Biometric sensor with enrolled fingerprint or face unlock Disables the biometric switch with a notice and keeps PIN access active.
Google Drive backup Google Play Services (GMS) Disables Drive actions with a notice; encrypted local file backup remains available.
Background alerts Notification permission (POST_NOTIFICATIONS) and unrestricted background battery settings Requests permission on enable; aggressive vendor battery limits may delay periodic checks.
Home-screen widgets Launcher that supports pinning widgets Hides or disables the pin action when unsupported by the launcher.

One-time payment, no subscription

Pro is bought once on Google Play and does not renew: there is nothing to cancel and no future charges. If you reinstall the app with the same Google account, use Restore purchases. See the terms and the refund policy.