New On-call in Maple: schedules, escalation and silences.

Traces, logs, errors, replays, uptime, on‑call. One app. Your agent reads it too.

$39/month with 100 GB each of logs, traces and metrics, then $0.30 per GB. No per-host or per-seat fees. Source on GitHub under FSL-1.1, Apache 2.0 after two years.

Start free trial
Engraving of a Roman temple above an aqueduct
  • Self-host it on your own infrastructure

    Every line of Maple's source is on GitHub under FSL-1.1, and each release becomes Apache 2.0 after two years. The self-hosted version runs for one team; alerting and team accounts are on Cloud for now.

    Read the source
  • No proprietary agent

    Maple speaks plain OpenTelemetry. Point an OTLP exporter at it and keep the instrumentation you already wrote. If you leave, there is nothing to rip out.

    See instrumentation
  • Hosted in the US or the EU us eu

    Choose the EU when you sign up and your telemetry is stored and processed in Frankfurt. Same plan and price in both regions.

    About EU hosting
  • $39/month, 14-day free trial

    A card is required to start the trial and nothing is charged until it ends. No per-seat, per-host or per-query fees at any volume.

    See what it costs
Health, latency and errors across your system.

Read the trace.

Every request as a span tree, attributes intact. Click a span and see what it carried.

POST /checkout 8 spans · 203ms
DurationStatus
/checkoutapi-gateway
203ms200
validateTokenauth-service
24.0msOk
createOrderorder-service
111msOk
validateInventory
20.0msOk
reserveItems
16.0msOk
processPaymentpayment-svc
57.0msOk
chargeCard
45.0msOk
sendConfirmationorder-service
30.0msOk

Search the logs.

Log Management

Structured logs straight from OTLP. Filter by severity, service or any attribute, then jump to the trace.

/logs production
TimeLevelMessage
INFOPOST /api/checkout 402
INFOorder canceled order_id=ord_8421 reason=payment_declined
WARNfalling back to authorize-only flow
ERRORcard declined: insufficient_funds
INFOcharging card amount=129.00 currency=EUR
DEBUGevaluate flag fast_checkout=false cohort=A2
INFOreserved 3 items sku=SKU-9081,SKU-1142,SKU-0077
WARNstock low sku=SKU-9081 qty=3
INFOtoken verified user=u_4109 scope=checkout
DEBUGjwks cache hit kid=2026-05
INFOPOST /api/checkout received
INFOGET /api/cart 200
INFOloaded cart user=u_4109 items=3
INFOGET /api/inventory 200
DEBUGreschedule retry_at=12:04:25 job=sync-inventory
INFOPOST /api/feedback 201
16 logs

Replay the session.

Browser Sessions

Every click, route, console line and failed request. Replay and spans share one session id.

A browser session replay showing the recorded page and captured network requests

From a 3 AM page to the line that caused it.

Four surfaces, one trace id. Every step below opens from the one before it.

  1. alert trace 4f9ac21e…8b7d40f2

    The alert arrives with context.

    Error rate on api-gateway crosses 5%. The alert carries the service, the threshold it broke and the spans that failed. It reaches you in Slack, PagerDuty, email or a webhook.

    An alert rule with its threshold, evaluation window and firing history
  2. trace trace 4f9ac21e…8b7d40f2

    The trace opens on the failing span.

    The failing request opens as a span tree. Three Stripe retries sit at the bottom, red, each timing out at exactly 1.75 seconds.

    A span selected in the waterfall with its full attribute set open alongside
  3. logs trace 4f9ac21e…8b7d40f2

    The logs are already correlated.

    Jump from the failing span to its logs on the trace id. Retry exhaustion, connection pool at 20/20. The confirmation is on the same screen.

    A log detail with linked trace and span IDs, correlated events and full resource attributes
  4. agent trace 4f9ac21e…8b7d40f2

    Or hand it to an agent.

    Point Claude, Cursor or any MCP client at Maple. It lists services, searches traces, reads the source and proposes the fix.

    An MCP client walking the same incident: five tool calls from list_services to propose_fix, ending in a patch
Engraving of a long Roman aqueduct crossing open country

I have no idea how [we] lived without Maple for so long.

Brian Anglin, CTO, Superwall
p99 across services, down from tens of seconds
<1s
traces, logs, and metrics a month
80B+
fewer errors
90%+
saved on compute every month
$1000s

Everything else you'd have bought separately.

Metrics and dashboards, a service map, error tracking, alerting and on-call, Kubernetes monitoring, web analytics, product events, uptime monitoring and GitHub checks, all on the instrumentation you already send.

A service map that builds itself

Live request flow between your services. When one fails, you see what it takes down with it.

service map · last 60s DEMO
2.8k req/s edge req/s 12.4k err% 0.1% avg 8ms checkout-api req/s 4.2k err% 0.3% avg 24ms cart-svc req/s 8.1k err% 0.1% avg 18ms auth-svc req/s 4.2k err% 0.0% avg 6ms pricing req/s 2.8k err% 2.4% avg 142ms payments req/s 1.1k err% 0.5% avg 38ms redis DB calls/s 18.2k avg 2ms postgres DB calls/s 6.4k avg 14ms
8 services · 8 edges · 60s window pricing-svc · p99 +18% vs 1h ago

Charts and traces that always agree

Request rate, error rate, latency percentiles. Build the board yourself, or ask the agent to build it.

metrics · http.server.request.durationDEMO
p5027 ms
p95175 ms
p99328 ms
checkout-api · 1m bucketssame spans as the trace view

142 errors, one issue

Errors grouped by type across every service, with trends over time, affected services and sample traces attached.

errors · issueDEMO
UpstreamErrorpayment-svc returned 500
events142
issues1
services3
12:10:00payment-svctrace 030d8d6f
12:09:57checkout-apitrace 109fa095
12:09:54order-workertrace 2c0d69e3
12:09:52payment-svctrace f5df28a5

Know why an alert fired, or why it didn't

Alert on error rate, latency, Apdex, throughput, any query-builder chart or raw SQL. Escalation chains and on-call schedules decide who is notified next. Routed to Slack, Microsoft Teams, Google Chat, Discord, Telegram, PagerDuty, incident.io, email or any webhook.

alerts · rule high-error-rate DEMO
rule critical
signalerror_rate
conditionerror_rate > 5%
window5 min · 2 consecutive
group byservice.name
5%
OK FIRING breaches 0/2 1/2 2/2 · eval every 60s
#alerts via slack
Maple
Maple APP 14:32
High error rate — checkout
error_rate at 8.4% · threshold 5% · window 5 min
service checkout · env production · severity critical
View in Maple Chat
delivered · 200 · attempt 1/5
slackdiscordtelegrampagerdutyemailwebhookhazel
evaluated every 60s · auto-resolve after 2 healthy

Web analytics from the sessions you already record.

Visitors, page views, referrers, countries, devices. No second script, no second vendor. MapleBrowser.init records the session. Analytics, replay and the trace all read from it. Crawlers are filtered out and the excluded count is shown. Click a referrer or a country and the page narrows to it. Every session links to its trace.

web analytics · last 7 daysDEMO
Unique visitors29,419
Page views122,373
Pages / session3.5
Online now34
Unique visitorsPage views

The pod behind the slow span.

Install a Helm chart and cluster metrics start flowing.

pod heatmap · 192 pods · cluster cpu 44% DEMO
PODS 192
RUNNING 184
DEGRADED 6
FAILING 2
HOT >78% 11
default default 72
observability obs 32
kube-system kube-sys 40
ingress-nginx ingress 24
data data 24
CPU
0% 100% degraded failing

Funnels on the telemetry you already send.

Count signups, checkouts and upgrades from the browser, your backend or a span you already emit.

funnel · pricing → paid conversion rate 6.4% DEMO
count by person window 24h range 30d
$pageview /pricing 100% 8,412 entered
signup_​started 35.2% 2,961 reached 35.2%
35.2% 64.8%
signup_​completed 26.0% 2,190 reached 74.0%
74.0% 26.0%
plan_​started 6.4% 538 reached 24.6%
24.6% 75.4%
signup_completed → plan_started
→converted538
↓dropped1,652
median time2h 14m
went next/docs/billing · 41%
  • plan_started trace trace 4f9ac21e
  • signup_completed server user u_8124
  • $pageview browser /pricing

The trace behind the failed check.

Maple checks your sites, APIs, ports and cron jobs around the clock.

uptime · 5 monitors DEMO
status monitor checks p95 30d
DOWN checkout-api GET /health · every 30s 212 ms 99.71%
UP auth GET /v1/session · every 60s 84 ms 99.98%
UP postgres-primary TCP db.internal:5432 · every 30s 3 ms 100%
UP acme.com MX DNS MX acme.com · every 5m 21 ms 100%
UP nightly-backup cron 0 3 * * * · grace 15m n/a 100%
incident opened 14:32:07
checkout-api is down
Status 503, expected 2xx · failed 2 checks in a row
trace 7c1e40a9…e21f service checkout sent to #oncall
failed checks 2/2 · closes when it passes again

Your pull requests, checked against production.

Install the GitHub App and Maple follows each change from the pull request to the hour after it deploys.

#1531 AI tab and email preferences with a local send schedule
6/10 needs attention

The AI key panel now loses its only Remove path for a lapsed add-on.

Warning correctness
Lapsed add-on leaves the saved OpenRouter key with no way to remove it ai-provider-section.tsx:179-181
migration Safe on production data
WebAnalyticsDigestService.ts 235 calls/day

Your data, your instrumentation, your bill.

OpenTelemetry in, no proprietary agent, and the source on GitHub the whole time. You can leave with everything you brought.

Maple on your laptop, as one binary.

One maple executable runs OTLP ingest, an embedded database, the query API and the dashboard on 127.0.0.1. No account, and your telemetry stays on your machine.

$ brew install Makisuo/tap/maple

or curl -fsSL https://maple.dev/cli/install | sh

More on Maple Local →

local · 127.0.0.1:4318 DEMO
maple start
🍁 maple · local mode
listening on http://127.0.0.1:4318
OTLP/HTTP POST /v1/{traces,logs,metrics}
query POST /local/query
dashboard https://local.maple.dev
· bundled UI: pass --offline
data ~/.maple/data
pid 48213 · stop with maple stop
one binary · one port telemetry stays on 127.0.0.1

Works with every language OpenTelemetry supports

  • Node.js
  • Next.js
  • Python
  • Go
  • Effect
  • OpenTelemetry
  • Kubernetes
  • Cloudflare
  • Prometheus

Let an agent run the investigation.

This is the sequence of tool calls an MCP client makes against Maple, from a service list to a proposed patch.

Reads the code behind the trace
The agent pulls the source file behind a span, so the fix it proposes cites your actual code.
Writes back
Claim an issue, set its severity, attach a fix. The agent changes state in Maple as well as reading it.
OAuth, no plugin
Point any MCP client at the endpoint and sign in. Open protocol, nothing to install.
mcp · agent session DEMO
> ask maple
list services that errored in the last 15m
→ list_services { window: "15m", filter: "errors > 0" }
← 3 services
paymentserrors142p992.4s recoserrors38p99840ms shippingerrors6p99320ms
> investigate payments
→ find_errors { service: "payments", window: "15m" }
← 1 issue
stripe.charge declined  ·  142 events  ·  98 users
→ error_detail { issue_id: "iss_8a91" }
← root cause: idempotency key collision
sample trace: trace_9f3c · order 8421
→ propose_fix { issue: "iss_8a91", fix: "scope idempotency_key by user_id" }
← fix attached. opened pr#412.
> 
4 tool calls · 4.2s https://api.maple.dev/mcp

One plan, billed by the gigabyte.

100 GB each of logs, traces and metrics included, then a flat rate. Nothing metered by host, seat or query.

Startup 14-day trial
$39 /month

Every feature is included.

$0
per host
$0
per seat
$0
per query
Start 14-day free trial

Free for 14 days · Cancel anytime · Card required to start

Hosted in the US or the EU (Frankfurt) at the same price. EU hosting

Included every month
Logs
100 GB
then $0.30 / GB
Traces
100 GB
then $0.30 / GB
Metrics
100 GB
then $0.30 / GB
Browser Sessions
5,000
then $0.002 / session
Product Events
1M
then $5.00 / 1M events
AI usage
$10
then billed per model, by tokens used

30 days of traces and logs, 90 of metrics · Private support channel · Unlimited dashboards · Advanced alerting · MCP server · AI error triaging · Full API access · Six-year audit log

Estimate your bill →
Enterprise

Higher volume, custom retention, priority support, SSO (SAML/OIDC) with SCIM, and your own cloud account.

The bill, line by line.

Each vendor's public list pricing next to Maple's.

Base pricePer-host feePer-seat feeIngest pricing Compare
Maple $39 / moNoneNone$0.30 / GB per signal
Datadog None$15+ / host / moNone$0.10 / GB + $1.70 / M indexed
New Relic NoneNone$10 first, then $99–349 / full-platform user / mo$0.40 / GB
Grafana Cloud $19 / moNone$8+ / active-user / mo$0.55 / GB
Dash0 NoneNoneNone$0.60 / M spans or logs · $0.20 / M metric points
SigNoz $49 / mo, incl. $49 usageNoneNone$0.30 / GB · $0.10 / M samples
Axiom $25 / moNoneNone$0.12 / GB + $0.20 / GB-hr
Better Stack None · bundles from $30None$29–34 / on-call responder / mo$0.10 / GB + $0.05 / GB-mo stored (EU)
Honeycomb $150 / mo (Pro)NoneNone$3.00 / M events (spans, logs)

Competitor cells summarize public list prices, checked September 2026 (base prices and free plans, November 2026). Check each vendor's pricing page for current numbers. Maple's rate is the published per-GB rate, billed per signal. Use the calculator to project your monthly cost.

Common questions

What is Maple?
Maple is an observability platform for traces, logs, metrics, errors, session replay, uptime checks and on-call, built on OpenTelemetry and backed by ClickHouse. You collect, query and visualize telemetry from your systems in one app, or let an AI agent do it for you over MCP.
Is Maple open source?
Maple is source available. The code is on GitHub under the Functional Source License (FSL-1.1): you can read every line, fork it and self-host it, but not offer it as a competing commercial service. Each release becomes Apache 2.0 two years after publication.
Is Maple OpenTelemetry-native?
Yes. Maple is built on OpenTelemetry, the open, vendor-neutral standard for telemetry, so there are no proprietary agents. If you already emit OpenTelemetry data, you can point it at Maple without re-instrumenting your code.
How is Maple priced?
The Startup plan is $39/month with 100 GB each of logs, traces and metrics included, then a flat $0.30 per GB. No per-host or per-seat fees. Enterprise plans add custom volume and retention.
Does Maple work with AI agents?
Yes. Maple runs an MCP (Model Context Protocol) server, so compatible AI agents can list services, search traces, find errors and propose fixes directly against your telemetry.
Can I self-host Maple?
Yes. Run the Maple container on your own server with a persistent volume. It bundles ingest, the dashboard and its own storage, for one team and without alerting. Maple Cloud is there for teams that need alerting, team accounts, or would rather not operate it.
Where is my data stored?
Maple Cloud runs in two regions: the United States and the European Union (Frankfurt, Germany). You choose one when you create your organization, and all of its telemetry is stored and processed there. The region cannot be changed later. Pricing is the same in both. To keep the data inside your own network, self-host Maple.
Engraving of a Roman triumphal arch on a hill

Try it on your own traces.

14-day free trial, card required, no charge until it ends. Or self-host, or run maple start locally with no account at all.