Toggled

Early access · launching soon · no card required

Feature flags platform for feature flag management, rollouts and A/B testing

Ship behind a flag, roll out 1% to 100%, run A/B tests, and kill a bad release in seconds. Flat pricing, unlimited seats.

Try the live demo below
checkout.new-flow ON

Rollout

37% of users

1%50%100%

Targeting rules

Add a rule and only matching users stay in the rollout.

Reverted in 0.9s (illustrative)

Who sees it, live

37 of 100 users see New checkout

new checkout old checkout

Experiment: checkout conversion

Illustrative sample data

A · old 3.1%
B · new 4.6%

Variant B converts +48% better in this sample. Roll it out with the slider above.

Turn the flag on to start splitting traffic between A and B.

A real flag console, running in your browser. Flip it, drag it, kill it.

Early access, launching soon No card required Your data stays yours SDKs for React, Node, Python, Go

01 · Why flags

Deploying code and releasing features are two different decisions

Tie them together and every launch is all-or-nothing. Separate them with a feature toggle and shipping becomes a dial you turn, not a button you pray over.

Ship faster

Merge to main behind a flag. No long-lived branches, no release-night freeze, no waiting for the quarterly train.

Roll out safely

Go 1%, then 10%, then 100% with one slider. Problems surface while the blast radius is a handful of users, not everyone.

Measure impact

Every rollout is already a split. Attach a metric and you know whether the feature helped before you commit to it.

Recover in seconds

The kill switch flips a bad release off for every user in about 3 seconds. No redeploy, no rollback window, no 2am deploy.

02 · The flag lifecycle

Five steps, left to right, no redeploys in between

See the full walkthrough
01

Create the flag

checkout.new-flow

Wrap the new code path in a flag with one SDK call. Deploy whenever you like; nobody sees it yet.

02

Target who

country = US · plan = pro

Point the flag at internal users, a beta cohort, a region or a plan. Rules evaluate locally in your SDK.

03

Roll out 1% to 100%

rollout: 37%

Drag the slider. Consistent hashing keeps every user in their cohort while you climb from a canary to everyone.

04

Measure

B +48% (sample)

The rollout is already an experiment. Attach a conversion or error metric and read the impact per variant.

05

Kill switch

reverted in 0.9s

Something regressed? One click turns it off for every user in about 3 seconds. The deploy stays where it is.

Swipe sideways through the lifecycle

03 · Hands on

Go on, break the checkout. You can’t.

The console at the top of this page is a real flag evaluator running in your browser. It is exactly how Toggled feels during an incident: calm, instant, reversible.

  • Flip the toggle and watch 100 users switch checkout flows in real time.
  • Drag the rollout to 8% to see what a canary release looks like: same dots, same cohort, every time.
  • Add country = US and the rollout instantly narrows to matching users only.
  • Hit the kill switch and everyone is back on the old flow before your pager finishes buzzing.
Back to the console

What your team gets

  • flags Unlimited flags and environments, named and searchable, with full change history.
  • sdks SDKs for React, Next.js, Node.js, Python, Go, Java, Rails and mobile.
  • safe Fail-safe defaults: if Toggled is unreachable, SDKs serve the last-known value. Your app never waits on us.
  • team Unlimited seats on every plan. Flags are infrastructure, not a per-seat line item.

experiment: onboarding.v2

Illustrative sample data

Variant A · current 18.2%
Variant B · new flow 23.9%

Confidence: 97%

Roll B out to 100%

04 · Experiments

A/B testing on the flags you already ship with

Other tools make experimentation a separate product with a separate bill. In Toggled, any flag becomes an experiment: split traffic between variants, attach a conversion or guardrail metric, and read the result in the same console. When B wins, the rollout slider is right there.

Read how it works in A/B testing software built on feature flags, or start with a gradual rollout and add the metric later.

05 · Pricing

Flat per team. Unlimited seats. The meter is gone.

Incumbents meter you: Unleash by the seat, LaunchDarkly by the service connection and by client-side monthly active users, Statsig by the event. Every one of those moves when you grow. Toggled is planned as flat team pricing: every engineer gets an account, every feature gets a flag, and finance never asks why the bill doubled. It is the LaunchDarkly alternative for teams that plan to grow, and we put the whole field side by side in best feature flag tools.

Launch

$39/mo

billed yearly · $49 monthly

  • Unlimited seats
  • Unlimited flags, 2 environments
  • 100k monthly contexts
  • Basic A/B testing
  • Kill switch
Most popular

Team

$119/mo

billed yearly · $149 monthly

  • Unlimited seats
  • Unlimited environments
  • 1M monthly contexts
  • Full experimentation
  • Audit log + roles

Scale

$399/mo

billed yearly · $499 monthly

  • Unlimited seats
  • 10M monthly contexts
  • Advanced experimentation
  • SSO + advanced RBAC
  • Uptime SLA + SOC 2 report

Enterprise

Custom

Talk to sales

  • Unlimited seats
  • Self-host or on-prem
  • SAML + SCIM provisioning
  • Custom retention + DPA
  • Dedicated support + SLA

Planned pricing for launch; early-access members lock it in. Full details on the pricing page.

06 · For engineering leaders

One prevented bad release pays for a year

Flags are risk infrastructure. The math that matters to a VP of Engineering is not the feature list, it is the incident you did not have and the per-seat bill you did not sign.

Rollback time

45 min → ~3 s

A redeploy becomes a kill switch flip. Incidents end in the console, not the war room.

Blast radius

100% → 1%

A canary deployment catches the regression while it touches a handful of sessions.

Cost per added engineer

$0

Flat team pricing. Hiring five engineers changes your flag bill by nothing at all.

Enterprise signals, stated honestly

SSO / SAML / SCIM Advanced RBAC Audit logs + retention Uptime SLA Self-host / on-prem SOC 2 planned DPA + invoicing

SDKs fail safe by design: flags serve their last-known value if Toggled is unreachable, so our outage is never yours. SOC 2 is planned, not yet certified; we say so because you would find out anyway.

07 · Questions

The honest answers first

All questions answered
Is Toggled live yet?

We are in early access. Join the list and we will email you when your spot opens. No card required, and we will not pretend otherwise with a fake countdown.

What will it cost?

Flat per-team pricing with unlimited seats: Launch $49, Team $149, Scale $499 per month, Enterprise custom. No per-seat meter. Early-access members lock in launch pricing.

How is this different from LaunchDarkly or Split?

Same capability class: flags, targeting, gradual rollouts and A/B testing. The difference is the bill. LaunchDarkly meters service connections and client-side monthly active users, so it tracks your pods and your traffic. Toggled is planned flat per team.

If Toggled goes down, do my flags break my app?

No. SDKs evaluate locally from a streamed ruleset and fall back to the last-known value. Your app never blocks on our API, and the kill switch still works during incidents.

Do I need a credit card to join?

No. Early access is a signup, not a purchase. You will only ever pay when the product launches and you choose a plan.

Early access · launching soon

Ship your next release behind a flag

Join the early-access list and be first in line when spots open. Flat pricing, unlimited seats, no card required.

No card required · your data stays yours

Keep exploring

More feature flag resources