Azure Monitor Pricing, Honestly: The Per-GB Trap (2026)

The Bill I Couldn’t Find

It was 2022, the same stretch I’ve confessed to burning two weeks opening AWS and Azure tabs I had no business opening. I had a Next.js app with eleven users, nine of whom were me, and I’d decided that a Real Company monitors its infrastructure. So I clicked around the Azure portal until I found Azure Monitor, wired up a Log Analytics workspace because a tutorial told me to, and admired the graphs for an afternoon.

A few weeks later I went looking for what it was costing me, and I genuinely could not find a single “Azure Monitor” line on the bill. There wasn’t one. Instead the cost was smeared across something called Log Analytics data ingestion, priced by the gigabyte, plus a little for alert rules, plus a retention charge I didn’t understand. My monitoring didn’t have a price. It had a meter, and the meter was measuring how much I logged.

That’s the whole thing you need to understand about Azure Monitor pricing, and almost nobody says it plainly: you don’t pay for the tool. You pay for the data you shove into it. It’s the exact same disease as CloudWatch on the AWS side — a dozen small meters instead of one honest number — just wearing a Microsoft badge.

---

What Azure Monitor Actually Is (One Name, Several Meters)

Azure Monitor is Azure’s native observability service. Like most Microsoft things it’s sold as one product but is really a cluster of separate jobs, each billed on its own meter. If the underlying words are fuzzy — what a metric is versus a log versus a trace, what you should never write to disk — I untangled that in monitoring and logging, explained, and I won’t re-run it here. For pricing, here’s the map that matters:

• Platform metrics and the Activity log — the CPU, memory, and “who changed what” data Azure collects automatically. These are free. Genuinely, always. This is the part that makes it feel free.
• Log Analytics — the workspace where all your logs and custom data land so you can query them. This is where the money is. Nearly your entire Azure Monitor bill is ingestion into this one thing.
• Alerts — rules that watch a metric or a log query and fire when something’s wrong. Cheap per rule, and they add up.
• Notifications — the actual text, email, or webhook when an alert fires. Billed separately from the alert itself.
• Application Insights — the app-level APM layer (page loads, request traces, exceptions). In 2026 it’s workspace-based, meaning its data flows into Log Analytics and is billed as… Log Analytics ingestion. Same meter, different door.

So five doors, one room where the money burns. Let me walk the numbers, because I checked every one of these against Microsoft’s own pricing page before writing it down — a stale figure would make this whole post sloppy.

---

The Number That Is Your Whole Bill: Per-GB Ingestion

Here’s the one to tattoo on your wrist. As of 2026, in the default US regions, Log Analytics Analytics Logs cost about $2.30 per GB ingested, after the first 5 GB per month, which is free. (It drifts a bit by region — some run closer to $2.76 — so check yours, but $2.30 is the anchor.)

Read that again. You pay by the gigabyte of logs you send in. Not stored. Not queried. Ingested. The moment your app produces a log line and ships it to the workspace, the meter clicks.

Five free gigabytes sounds like plenty until you meet a real app. Leave debug logging on in one chatty service. Let a bot hammer an endpoint that logs every request. Turn on verbose diagnostic settings for a database “just to see.” Any of those can dump tens of gigabytes in a day, and the meter does not care that 95% of it was noise nobody will ever read. This is the exact way founders get murdered on Datadog’s per-GB logs too — usage-based pricing punishes the person who isn’t watching, and a new founder is never watching the log meter.

Microsoft does give you cheaper lanes if you plan ahead:

• Basic Logs: ~$0.50/GB — for high-volume, low-value logs you rarely query. Cheaper to ingest, but querying them costs extra and retention is limited.
• Auxiliary Logs: ~$0.05/GB — the bargain bin, for verbose data you almost never touch.

The catch is you have to classify your log streams into the right tier ahead of time. Nobody bootstrapping does this. You send everything to Analytics Logs at $2.30/GB because that’s the default, and you find out about Basic and Auxiliary after the first invoice makes you go reading.

---

Retention: The Rent You Pay to Keep Your Own Logs

Ingestion gets you in the door. Retention is the rent.

Your data sits in the interactive (hot) tier free for 31 days. Keep it longer — and you will, because compliance, or debugging a bug that only shows up monthly, or just not knowing you could delete it — and it’s about $0.10 per GB per month to hold in the interactive tier, up to two years. There’s a long-term / archive tier at roughly $0.02 per GB per month if you drop the data somewhere cold and accept that pulling it back for a search costs more.

Put ingestion and retention together and you see the trap clearly. You pay $2.30 to bring a gigabyte in, then you pay every single month to keep it, and the default is to keep everything. A founder who logs 30 GB/month and never sets a retention policy is signing up for a bill that grows on its own — more data, more months, more rent — while the product still has thirty users. This is the same “storage that quietly bleeds you” pattern I hit with Azure Blob egress fees: the sticker price on the first action is never the whole story.

---

Alerts and Notifications: Cheap Alone, Not in a Pile

The useful part of any monitoring tool is the alert — the smoke detector that texts you when the site’s on fire. Azure prices these à la carte too.

Metric alerts cost about $0.10 per monitored time series, per month, and you get 10 free. A “time series” is one metric on one dimension — CPU on this VM, error rate on that endpoint. Sounds trivial. But teams instrument everything, split by every dimension, and “just ten alerts” quietly becomes 200 time series and a line item you have to explain.

Log alerts — rules that run a Log Analytics query on a schedule — run about $0.50 for the first time series and $0.05 for each additional one, monthly, and the tighter you set the evaluation frequency the more they cost. Notifications are billed on top, separately, by channel: emails and push notifications are cheap with a monthly free allowance, webhooks are near-free at low volume, and SMS is metered per message and varies by country — which is the one that surprises people who wire up “text me on every alert” and then have a noisy week.

None of these is outrageous on its own. That’s the whole point. Azure Monitor doesn’t have a price — it’s the sum of a hundred small decisions you made without noticing, and each one clicks its own meter.

---

Commitment Tiers: A Discount You Have to Earn by Burning

Once you’re ingesting real volume, Azure offers commitment tiers — you promise to buy a fixed amount of ingestion per day, and the per-GB rate drops. As of 2026 it starts at 100 GB/day for about $1.96/GB (~15% off), 200 GB/day for ~$1.84/GB (~20% off), and 500 GB/day for ~$1.61/GB (~30% off). Overage above your commitment bills at that same discounted rate, and you’re locked in for a minimum of 31 days.

I’m including these for completeness, but let me be blunt about who they’re for. To break even on the smallest tier you have to be ingesting 100 gigabytes a day. That is an enterprise with real traffic and a real logging problem. If you’re reading a blog by a guy who paid ~$90/month with $0 MRR, the commitment tiers are not your discount — they’re the sign on the door telling you this whole product was built for someone else. A bootstrapper who “commits” to 100 GB/day to save 15% has committed to a bill that would have made me physically ill in 2022.

---

The Broken Engineer Verdict

Same fork in the road as the AWS monitoring post, same one question: are you already on Azure?

If you are — your app runs on Azure App Service, Azure SQL, AKS, the whole family — then use Azure Monitor. Don’t bolt a third-party tool onto an Azure app just to dodge it; now you’re paying twice and shipping your logs across the internet for nothing. It’s right there, it’s integrated, platform metrics are free, and for uptime and a few alerts it does the job. Just treat every meter like it’s live: turn off DEBUG logging in production, classify high-volume noise into Basic or Auxiliary logs before the first surprise invoice, set a retention policy instead of hoarding everything for two years, and don’t create 200 metric alerts for a product with 30 users. Azure Monitor is fine. It’s just fiddly, and it never stops counting your gigabytes.

If you’re not on Azure — and if you read my take on Azure vs AWS, you know I think most founders shouldn’t be on either — then you should never open Azure Monitor in your life. A Log Analytics bill for someone running on Vercel, Railway, and Neon is a bill you invented. The three things you actually need from zero to a hundred users — is it up, did it crash, let me read the logs — are all free elsewhere, and I laid that $0 stack out in full in the Datadog post: Better Stack or UptimeRobot for uptime, Sentry (or PostHog if you already run it) for errors, and Grafana Cloud’s absurdly generous free tier for logs and metrics if you want graphs. Same jobs. No per-GB meter.

The mistake was never Azure Monitor. It’s a good tool for Azure shops. The mistake is dragging yourself onto Azure because you wanted the monitoring, or paying a metered per-gigabyte price to watch a product that hasn’t earned a rupee.

---

The Bottom Line

Azure Monitor pricing confused me for weeks in 2022 for one reason: I kept looking for the price of the tool, and the tool doesn’t have one. Its price is your log volume. Ingest 5 GB, pay nothing. Ingest 500 GB of debug noise you’ll never read, pay for all of it — twice, once to bring it in and again every month to keep it. That’s not a hidden fee. It’s the entire business model, and it’s the same model as CloudWatch and Datadog wearing different logos.

Monitoring should be a smoke detector, not a warehouse you rent by the gigabyte. If you already live on Azure, Azure Monitor is the detector in the wall — use it, and watch your ingestion like it’s live, because it is. If you don’t, you already have a smoke detector, and it’s free. Don’t go install a metered one next to it just because it has the Azure logo on the box. (And if you’re still adding up every line of your monthly burn, I broke down what it actually costs to run a SaaS — observability doesn’t have to be on that list at all.)

This is the Broken Engineer Guide — I over-engineer everything, fail at business, and hand you the receipts so you don’t repeat my 2022 portal-clicking. Go watch your uptime for free, then go build something worth monitoring.