See and govern your AI agents.
In real time.

As agents move from experiments to mission-critical work, you need one place to see not just what every agent is doing, but how much it is costing, what is stuck, and what needs your approval, in real time, across every runtime, model, and workflow.

End-to-end encrypted AES-256-GCM client-side encryption on every tier. Your agents' conversations and activity content are encrypted on your machine before they leave, so we store ciphertext we cannot read, even if subpoenaed. Simple totals like cost and health sync securely to power the dashboard. Annual: 2 months free + free desk device Pay annually and get 2 months free (10x the monthly price). Every annual plan also includes the $149 ClawMetry desk device, free. See the device →
Choose how to run ClawMetry
Free
$0 forever
1 node · OpenClaw + NeMo · no account
Watch one AI agent for free, forever. No account needed. You get the live dashboard, cost tracking, and encryption right out of the box.
Included
  • ✓ OpenClaw runtime + NeMo Guardrails
  • ✓ 1 node monitoring
  • ✓ Brain feed (24 hours)
  • ✓ Token tracking (7 days)
  • ✓ E2E encryption + secret redaction
  • ✓ Tamper-evident audit log
  • ✓ Community support
Get Started Free
Best value
Starter
$19$9 / node / month
or $90/node/year · 2 months free
✓ Includes a free $149 desk device
7-day free trial · no credit card
Run all your AI agents on one dashboard. See what each one is doing, watch Claude Code, Cursor, Aider and 9 more side by side, get alerts when costs spike, and approve risky actions before they run.
Everything in Free, plus
  • ✓ Multi-runtime: Claude Code, Codex, Cursor, Aider, Goose, OpenCode, Qwen, Hermes, PicoClaw, NanoClaw
  • ✓ Add unlimited nodes ($9/mo each) + fleet view
  • ✓ Cloud sync (E2E encrypted snapshot)
  • ✓ Brain feed (7 days)
  • ✓ Token + cost split (30 days)
  • ✓ All 21 channel adapters (Telegram, Slack, WhatsApp, Discord, IRC, ...)
  • ✓ Approval queue
  • ✓ Budget limits + alerts
  • ✓ Anomaly detection (spend + error spikes)
  • ✓ Cost optimizer + model routing
  • ✓ Flow + per-runtime health timeline
  • ✓ Email support (24h response)
Get Started →

$9/node/month after trial · cancel anytime

Recommended
Pro
$39$19 / node / month
or $190/node/year · 2 months free
✓ Includes a free $149 desk device
For teams running agents in production. Catch runaway loops and waste in real time, let agents fix their own mistakes, lock down what tools can do before they run, and send everything to your monitoring stack.
Everything in Starter, plus
  • ✓ Per-run waste flags (runaway, cold cache, bloated context)
  • ✓ Per-run compare (A vs B with deltas)
  • ✓ Error triage (resolve / mute known errors)
  • ✓ Self-Evolve findings + Fix-with-AI
  • ✓ Asset registry (skills, prompts, workflows)
  • ✓ Eval suite (LLM-as-judge scoring)
  • ✓ Tool catalog policy + pre-execution gate
  • ✓ OTel export (Datadog, Grafana, Honeycomb)
  • ✓ Custom webhooks + PagerDuty + OpsGenie
  • ✓ Custom runtime HTTP ingest API
  • ✓ 90-day event retention
  • ✓ Priority email (4h response)

Or choose Starter and upgrade anytime.

1 node = 1 OpenClaw machine  ·  All prices in USD

Enterprise
For 50+ nodes, regulated industries, and air-gapped deployments.
Everything in Pro, plus the controls and deployment options that pass procurement at banks, defense, healthcare, and infrastructure teams running AI agents in production. Annual contract, custom pricing.
A custom fleet desk device, built to your needs: watch your whole agent fleet at a glance, get notified the moment one needs attention, and run hands-off in auto-pilot, without juggling tabs. See the device for reference →
  • 🔐 SSO + SAML + OIDC
  • 👥 RBAC + teams
  • 📋 Tamper-evident audit log
  • 🔒 Secret redaction at ingest
  • 📡 SIEM export (Splunk, QRadar, Elastic)
  • 📡 OTel export
  • 🔒 On-prem + air-gapped license
  • 🌍 Custom data residency
  • ⚡ 99.9% SLA + dedicated Slack
  • ⏳ Custom retention
Security and procurement pack →
Book a 30 min call →

Or email enterprise@clawmetry.com. Reply within 1 business day.

Compare all features

Free Starter Pro Enterprise
Runtimes
OpenClaw runtime + NeMo Guardrails✓✓✓✓
Claude Code, Codex, Cursor, Aider, Goose, OpenCode, Qwen, Hermes, PicoClaw, NanoClaw·✓✓✓
Custom runtime HTTP ingest API··✓✓
Observability
Overview, Tool catalog, Sub-agents, Memory, Crons✓✓✓✓
Brain feed retention24h7d30dcustom
Token + cost retention7d30d90dcustom
Tracing, spans, waterfallbasic✓✓✓
Cost attribution + model attribution·✓✓✓
Per-runtime health timeline·✓✓✓
Analytics and intelligence
Per-run waste flags (runaway, cold cache, bloated context)··✓✓
Per-run compare (A vs B with deltas)··✓✓
Error triage (resolve / mute known errors)··✓✓
Self-Evolve findings + Fix-with-AI··✓✓
Asset registry (skills, prompts, workflows)··✓✓
Eval suite (LLM-as-judge scoring)··✓✓
Anomaly detection (spend + error spikes)·✓✓✓
Cost optimizer + model routing·✓✓✓
Governance and control
NeMo Guardrails surface✓✓✓✓
Approval queue (block tool calls by policy)·✓✓✓
Budget limits + alerts·✓✓✓
Custom alert webhooks + PagerDuty + OpsGenie··✓✓
Tool catalog policy + pre-execution gate··✓✓
Integrations
Channel adapters (Telegram, Slack, WhatsApp, Discord, ...)3all 21all 21all 21
OTel export (Datadog, Grafana, Honeycomb)··✓✓
SIEM export (Splunk, QRadar, Elastic)···✓
Cloud and multi-node
Number of nodes1unlimitedunlimitedunlimited
Cloud sync (E2E encrypted snapshot)·✓✓✓
Fleet view (multi-node dashboard)·✓✓✓
Self-hosted cloud dashboard··✓✓
Air-gapped on-prem deployment···✓
Security and compliance
E2E encryption (AES-256-GCM)✓✓✓✓
Secret redaction at ingest✓✓✓✓
Tamper-evident hash chain audit log✓✓✓✓
SSO (SAML, OIDC, Okta, Google, Azure AD)···✓
RBAC + teams + workspace scoping···✓
Custom data residency + DPA + sub-processor list···✓
Support and SLA
Community Discord + GitHub issues✓✓✓✓
Email support response·24h4hnamed CSM
Dedicated Slack with the founding team···✓
99.9% uptime SLA···✓

Common questions

What do I get for free?
The Free plan monitors 1 node forever with no account required. You get the overview dashboard, 24 hours of Brain feed history, 7 days of token tracking, and E2E encryption. It's everything you need to get started.
What's the difference between Starter and Pro?
Starter ($9/node/month) is for teams who want to observe AI agents across multiple runtimes (Claude Code, Codex, Cursor, Aider, Goose, OpenCode, Qwen, Hermes) in one dashboard, with unlimited nodes, cloud sync, all 21 channel adapters, and approval queue. Pro ($19/node/month) is for production teams that need to stop runaway loops and waste, evolve agents on their own findings, and enforce policies before tools fire. Pro adds Self-Evolve, Fix-with-AI, asset registry, eval suite, per-run waste flags, per-run compare, error triage, OTel export, custom webhooks, and a custom runtime HTTP ingest API.
Is my data safe?
Yes. Your agents' conversations and activity content are E2E encrypted on your machine before they ever leave - we use AES-256-GCM client-side encryption, and the server stores that content only as ciphertext we cannot read. Simple totals (cost, token counts, health status) sync securely in readable form so your dashboard works instantly.
What counts as a node?
A node is one machine running ClawMetry's sync daemon. Free covers 1 node. Starter is $9/node/month and Pro is $19/node/month: 5 nodes on Starter is $45/month, 5 on Pro is $95/month. Enterprise is an annual contract with custom pricing.
How does the free trial work?
Starter comes with a 7-day free trial, no credit card required. You get full access to every Starter feature. If you don't upgrade after 7 days, you drop back to the Free plan, no charges, no surprises. Pro doesn't have a self-serve trial; book a 30 min call and we'll set up an evaluation.
What happens if I cancel?
Your data is retained for 90 days after cancellation, then permanently deleted. Your Free plan access continues to work - you never lose single-node monitoring.