Opsgenie is going away. Not “deprioritized,” not “in maintenance mode” — actually turned off. Atlassian stopped selling new Opsgenie accounts and trials on June 4, 2025, and the product reaches end of support on April 5, 2027. After that date, alerts stop routing, on-call schedules stop firing, and the data gets deleted. If your team still wakes up to an Opsgenie push at 3 a.m., you have a migration on your plate, and the clock is already running.
I’ve moved two teams off Opsgenie in the last year, and the thing that surprised me both times was how much of the work isn’t technical. Exporting schedules and rebuilding escalation policies is a weekend. Re-pointing forty integrations, retraining everyone on a new mobile app, and getting finance to approve a line item that didn’t exist before — that’s the part that eats a quarter. So this isn’t just “which opsgenie alternative has the best feature list.” It’s “which one can my team actually be live on before our next renewal, at a price I can defend.”
Here’s how PagerDuty, incident.io, Grafana Cloud IRM, Better Stack, and Squadcast stack up, and how I’d pick between them depending on the size and shape of your org.
The timeline, and why “just use Jira Service Management” isn’t the whole answer
The dates that matter:
- June 4, 2025 — no new Opsgenie accounts or trials. If you don’t already have one, you can’t get one.
- Around October 2025 — teams using the Opsgenie capabilities bundled inside Jira Service Management lost the standalone Opsgenie experience earlier than everyone else.
- April 5, 2027 — full end of support. The REST APIs keep responding until then, but after that the service is off and data is removed.
- 120 days after you start a migration — Opsgenie shuts down automatically. So once you flip the switch, you’ve committed.
Atlassian’s official answer is “migrate to Jira Service Management” (Compass gets mentioned for some service-catalog use cases, but JSM is the headline). And honestly, if you already run JSM for your service desk, that’s a reasonable default — the alerting, on-call, and incident features now live there, the migration tooling is built, and you’re not adding a vendor.
But a lot of Opsgenie users never touched the rest of Atlassian’s stack. They bought Opsgenie standalone because it was a clean, focused on-call tool that didn’t care what ticketing system you used. For those teams, “move to JSM” means adopting a whole service-management product to get the on-call piece back, plus JSM’s per-agent pricing on top. That’s not a migration, that’s a platform decision. If you’re in that camp, the field is wide open, and some of the alternatives below are genuinely better at the on-call job than Opsgenie ever was.
What to actually evaluate
Strip away the marketing and an on-call/incident tool is doing five things:
Schedules and escalation. Layered rotations, overrides, follow-the-sun, “if Priya doesn’t ack in 5 minutes, page the secondary, then the manager.” This is table stakes — every tool here does it — but the editing experience and how schedules behave around timezone changes and holidays varies a lot.
Alert ingestion, dedup, and routing. How many integrations ship out of the box, how good the noise reduction is (grouping, suppression, maintenance windows), and whether routing rules are sane to write.
Notification reliability. Mobile push, phone calls, SMS — and crucially, what happens when push fails. This is the one feature you can’t compromise on, and it’s also where the per-message costs hide.
Incident coordination. Slack/Teams-native incident channels, roles, timelines, status updates, retros. Opsgenie was always weak here; the newer tools treat it as the main event.
Status pages, integrations, and compliance. Public/private status pages, a real API and Terraform provider, SSO/SCIM, SOC 2, HIPAA if you need it.
Keep one number in your head while you read: count responders, not headcount. Almost every tool here charges per person who’s actually on a rotation or acknowledges alerts. Stakeholders who just want to watch a status page are usually free or cheap. A 60-person engineering org might only have 18 real responders.
PagerDuty — the closest like-for-like replacement
If you want the smallest possible behavioral change for your responders, PagerDuty is it. The schedule model, escalation policies, and alert-routing concepts map almost one-to-one onto how Opsgenie worked, the mobile app is the most battle-tested in the category, and the integration catalog is the biggest by a wide margin. There’s a documented migration path and tooling to bring schedules and services over.
Where PagerDuty has pulled ahead is event intelligence — the AIOps layer that groups related alerts, suppresses transient noise, and surfaces probable causes. On a busy platform team that’s the difference between one page and forty. The flip side: a lot of that lives in the higher tiers, and PagerDuty has a long history of “the feature you want is one plan up.”
Pricing, as of May 2026: a Free plan for up to 5 users, then list pricing in the rough neighborhood of $21 per user per month for Professional and ~$41 per user per month for Business (annual billing — published numbers vary a bit by source and you should expect a sales conversation), with a custom-priced Enterprise/Digital Operations tier above that. SMS and phone notifications beyond the included allotment are metered.
The honest read: PagerDuty is the safe choice and usually the most expensive one. If you’re a 200-person org with compliance requirements and a culture built around PagerDuty-style workflows, paying for it is defensible. If you’re 25 engineers, you’ll feel every seat.
incident.io — if your incidents already live in Slack
incident.io came at this from the other end. It started as a Slack-native incident-response tool — declare an incident with a slash command, it spins up a channel, assigns roles, builds the timeline, posts updates, drafts the retro — and then added On-call as a module. So instead of “alerting tool that bolts on incident features,” it’s “incident tool that added paging.” For a lot of modern teams that’s the right shape: the coordination layer is where you actually spend the painful hours of an outage, and incident.io is the best at it.
The trade-off is that On-call is a separate paid add-on. The base incident-response product runs roughly $19–$20 per user per month on the Team/Pro tiers, and On-call adds another $12–$20 per user per month on top, so a responder lands somewhere around $31–$45 per user per month depending on plan; Enterprise (think 300+ users) is custom and tends to land near $50. That’s PagerDuty-Business money or more — but you’re getting a meaningfully better incident-management experience for it, not just paging.
I’d pick incident.io if your team runs incidents in Slack today, your retros are inconsistent and you want that fixed, and you’re okay with the bill. I would not pick it purely as an Opsgenie paging replacement — that’s overpaying for the half you’d use.
Grafana Cloud IRM — the cheap path if you already run Grafana
If your monitoring stack is already Grafana and Prometheus, Grafana Cloud IRM (the merged product formerly known as Grafana OnCall plus Grafana Incident) deserves a hard look, because the economics are unusual. IRM is billed per monthly active IRM user — someone who’s on a schedule or escalation chain, or who acts on an alert group — and on the Grafana Cloud Free tier it’s genuinely free for up to 3 active IRM users. Past that you need a paid Grafana Cloud plan, and it scales per active user from there; the Enterprise tier carries a minimum commitment (on the order of $25k/year), which tells you who that tier is for.
What you give up: this is on-call and incident response that’s good enough, attached to a product whose center of gravity is dashboards and metrics. The escalation logic, mobile experience, and integration breadth aren’t at PagerDuty’s level, and the incident-coordination side is younger than incident.io’s. If you’re not already a Grafana shop, adopting Grafana Cloud just to get IRM is the same mistake as adopting JSM just to get on-call — you’re buying a platform for a feature.
But if you are a Grafana shop with a small team? It’s hard to argue with “the alert that fired the dashboard can page you, from the same console, for roughly nothing.” That’s a real option Opsgenie never gave you.
Better Stack and Squadcast — the budget end
For small teams, the per-seat math on PagerDuty and incident.io gets silly fast, and these two are where I’d look.
Better Stack bundles uptime monitoring, on-call, incident management, and status pages into one product, and it’s aggressive on price: the on-call/incident plan starts around $29 per responder per month on annual billing (~$34 monthly), and notably it includes unlimited phone-call alerts at that price — no per-call metering, which is the line item that quietly inflates PagerDuty bills. There’s a free tier with real (if small) allowances.
The catch is the à la carte add-ons: Slack/Teams incident workflows run about $9 per responder per month extra, additional status pages are billed separately, and advanced call routing is its own line. Price it with the add-ons you’ll actually use, not the headline number.
Squadcast is the closest thing to “Opsgenie, but cheaper and still focused.” Free plan for small teams, then Pro at ~$12 per user per month, Premium at ~$19, Enterprise at ~$26 — all per user. You get on-call, escalations, incident response, and (on Premium and up) status pages and some SRE-flavored extras like SLO tracking. SMS and voice past the included allowance are metered (~$0.10 per message for US/India/Brazil, ~$0.35 elsewhere), so factor that in if you lean on phone alerts. At Squadcast’s price point, a 15-person rotation costs less than a couple of PagerDuty seats.
Where both fall short: scale and ecosystem depth. The biggest integration catalogs, the most mature AIOps/event-intelligence, the enterprise procurement experience, the long list of compliance attestations — that’s still PagerDuty’s turf. If you’re a 5–30 person team that mostly needs reliable paging and clean schedules, you won’t miss any of it. If you’re 300 engineers across a dozen services with auditors asking questions, you might.
(Worth a quick name-check: Rootly and FireHydrant also pitch hard for Opsgenie refugees and lean toward the incident-coordination/process side like incident.io. If that’s your priority and you want a third quote, add one of them to the bake-off.)
Pricing at a glance (list pricing, as of May 2026)
| Tool | Free tier | Paid entry (per responder/mo) | Top published tier | Notes |
|---|---|---|---|---|
| PagerDuty | Up to 5 users | ~$21 (Professional, annual) | ~$41 (Business); Enterprise custom | Biggest integration catalog; best AIOps; SMS/calls metered above allotment |
| incident.io | Limited | ~$19–$20 base + ~$12–$20 On-call add-on | ~$50 Enterprise (custom) | Best incident coordination; On-call is a separate module |
| Grafana Cloud IRM | Yes — up to 3 active IRM users | Per active user on a paid Grafana Cloud plan | Enterprise ~$25k/yr minimum | Cheapest if you already run Grafana/Prometheus |
| Better Stack | Yes (small allowances) | Higher tiers + add-ons | Unlimited phone calls included; Slack/Teams workflows, extra status pages cost extra | |
| Squadcast | Yes (small teams) | ~$12 (Pro) | ~$26 (Enterprise) | Closest “Opsgenie but cheaper”; SMS/voice metered (~$0.10–$0.35/msg) |
Treat these as starting points for a quote, not gospel — every vendor in this space negotiates on committed seats, and published numbers drift. Always price the add-ons you’ll actually turn on, not the headline seat price.
A migration checklist that won’t blow up at 3 a.m.
The order matters more than the steps.
- Inventory first. List every Opsgenie integration (monitoring tools, ticketing, chat, webhooks, custom scripts), every schedule, every escalation policy, every team, and — this one bites people — every place an Opsgenie API key or webhook URL is hardcoded in CI, IaC, or a runbook.
- Pick on responders × price, then features. Multiply your real responder count by the per-seat price for each finalist. That number eliminates options faster than any feature spreadsheet, and it’s the number finance will ask for.
- Rebuild schedules and escalations by hand. The exporters help, but layered rotations, overrides, and timezone behavior deserve a human checking them. A schedule that’s “mostly right” pages the wrong person on a holiday.
- Run both systems in parallel for at least two weeks. Fan critical alerts to old and new simultaneously. You’re looking for the alert that doesn’t show up in the new tool — silent gaps, not loud ones.
- Cut over per team, not all at once. Move one team, let them live on it through a real incident, fix what’s wrong, then move the next.
- Then start the official 120-day clock. Remember Opsgenie shuts down automatically 120 days after you begin migration — so don’t “start” until you’re genuinely ready to finish.
- Aim to be fully live ~6 months before your renewal, not your shutdown date. That gives you a quarter of breathing room and a clean exit from any Opsgenie/Atlassian contract overlap.
So what should you actually do?
Solo dev or tiny startup: Squadcast’s free or Pro tier, or Grafana Cloud IRM if you already run Grafana. Don’t pay PagerDuty prices for a three-person rotation.
Small SaaS team (≈10–40 engineers): Squadcast or Better Stack. Better Stack if you also want uptime monitoring and status pages in one bill and value unlimited phone alerts; Squadcast if you want the cleanest pure on-call experience for the lowest per-seat cost.
Mid-size eng org with a real incident culture: incident.io, if your incidents already live in Slack and you want retros and coordination to actually get better — and you can stomach ~$31–$45 per responder. Otherwise PagerDuty.
Enterprise with compliance and a big integration footprint: PagerDuty. It’s the expensive answer and the safe one, and at that scale “safe” is worth real money. (If you’re already deep in Atlassian, JSM is the no-new-vendor alternative — just price the per-agent cost honestly.)
The mistake I’d avoid: drifting onto Jira Service Management by default just because Atlassian pointed there, without checking whether a focused on-call tool does the job better and cheaper for what you actually need. For plenty of teams it does. The forced migration is annoying, but it’s also a free excuse to fix the incident process you’ve been complaining about for two years.
Pick your two finalists, get both into a real on-call rotation for a week, and watch how they behave during an actual page. The tool that feels boring under pressure is the one you want.