Figment vs AWS: Mac Hosting Compared in 2026

Verdict: Figment's $95 M-series entry vs AWS Mac on-demand

AWS EC2 Mac runs on Dedicated Hosts billed per second with a 24-hour minimum allocation period per host. The current published US East (N. Virginia) on-demand rate for the M1 generation (mac2) is $0.65/hour, read directly from the AWS metered pricing feed at b0.p.awsstatic.com/.../dedicatedhost-ondemand.json on 2026-11-10 (publication timestamp 2026-11-10T17:38:52Z). The same feed lists mac-m4 at $1.23/hour, mac2-m2 at $0.878/hour and mac1 (Intel) at $1.083/hour.

Figment's M-series Mac is $95/month entry. That is the advertised entry price; the exact chip, RAM and disk are set when we respond to your configuration request, so treat the figure as an entry-price comparison, not a spec-for-spec match. If an AWS Mac runs always-on for a full month normalized at 730 hours, the M1 mac2 compute alone is $474.50, before EBS and egress. Figment's entry plan is lower on monthly outlay for steady, always-on work. AWS wins when you need to burst many Macs in parallel for a release window or sit the Mac inside an existing VPC alongside IAM, S3 and the rest of your account.

If you want Figment to price your exact configuration, start here:

Request your Figment M-series Mac →

AWS Mac hourly rates verified from the official pricing feed

The table below is the US East (N. Virginia) on-demand row from the primary AWS feed cited above. Other regions can differ; the AWS EC2 Dedicated Hosts pricing page is the same feed rendered as HTML and lets you switch regions.

Two caveats on that right-hand column:

• 730 h is a normalized month, not every actual invoice. A billing period with 744 hours (31 days) or 672 hours (28 days) will differ, and the 24-hour minimum applies per allocation, not per month—releasing and re-allocating a host within 24 hours still bills the full 24 hours for each allocation.
• Storage and network are extra. EBS volumes, snapshots and inter-region or internet egress are billed on top of the hourly compute line. The AWS EBS pricing page is the source for the gp3 and io2 rates in your region; treat the storage figure in any worksheet as a parameter you set (region rate × GB × month), not a fixed AWS-wide number.

The AWS EC2 Mac product page also states Savings Plans can reduce on-demand by up to 44% with a 3-year commitment. Applied to the M1 rate: 0.65 × 0.56 = $0.364/hour → $265.72 at 730 hours—still above Figment's $95 entry for steady-state workloads, and only when you accept the commitment term. Instance and Compute Savings Plans have different coverage rules; check the AWS Mac page for your specific commitment before quoting a number internally.

Break-even worksheet and a worked monthly cost

A single division tells you where AWS on-demand overtakes a $95 Figment entry on compute alone:

Break-even (hours/month) = Figment monthly price ÷ AWS hourly rate
    mac2    (M1):    95 ÷ 0.650 ≈ 146.2 h
    mac2-m2 (M2):    95 ÷ 0.878 ≈ 108.2 h
    mac1 (Intel):    95 ÷ 1.083 ≈  87.7 h
    mac-m4  (M4):    95 ÷ 1.230 ≈  77.2 h
    mac2-m2pro:      95 ÷ 1.560 ≈  60.9 h
    mac-m4pro:       95 ÷ 1.970 ≈  48.2 h

Plug your real numbers in. For always-on CI, one Mac, 40 builds/day on the M1 generation:

• AWS mac2 (M1) compute at 730 h: $474.50
• EBS gp3 for derived data and simulators (parameter; set from your region's per-GB-month rate × provisioned GB): $S
• Egress to TestFlight and artifact store (parameter; set from your monthly GB × region egress rate): $E
• AWS monthly total: $474.50 + S + E
• Figment M-series entry (always-on, flat): $95
• Figment Managed Build Mac (one M-series Mac, CI setup, backups, monitoring, 30 minutes monthly support): $249

The 24-hour minimum means a Mac that "runs CI a few times a day" is effectively always-on unless someone actively releases the Dedicated Host between jobs. At 24 h × $0.65 = $15.60 per allocation on M1, one allocation per day for 20 weekdays is $312/month—below steady-state M1 but above the Figment entry for the same month.

Hardware, access and toolchain fit

AWS EC2 Mac offers the nine generations in the table above, placed on Dedicated Hosts inside your VPC with access controlled by Security Groups, EBS, IAM and the rest of your AWS account.

Figment's $29 virtual Intel macOS plan is 2 vCPU, 8 GiB RAM, 80 GiB disk on Linux/KVM. There is no GPU acceleration and no audio; it is not Apple silicon. The verified access path is the browser console, which includes the VM's own desktop terminal for shell work; outbound CI agents on the VM can connect to your GitHub, Buildkite or GitLab service. Use the Intel VM for Intel-compatible Xcode releases. For Apple silicon toolchains, GPU work or visionOS, request an M-series Mac from $95/month, the Managed Build Mac at $249/month, or the Managed Mac Team at $999/month (two M-series Macs, two CI runners, team setup, backups, monitoring and one hour monthly support). Physical provisioning is request-based: chip, RAM, disk, term and access method are confirmed in the quote.

Toolchain references (primary, from Apple's Xcode system requirements):

• Xcode 27 requires macOS Tahoe 26.6 or later and runs on Apple silicon only—Intel Macs cannot run Xcode 27.
• Xcode 26 point releases pair with macOS Sequoia 15.6 through Tahoe 26.x; confirm the exact row for the release you need.
• visionOS development requires a Mac with Apple silicon.

macOS Tahoe itself supports several Intel Macs (Apple's Tahoe compatibility list includes MacBook Pro 2019/2020, iMac 2020 and Mac Pro 2019); the Apple-silicon-only restriction applies to Xcode 27, not the operating system. If your build matrix pins to Xcode 26 or earlier, an Intel target remains viable; for Xcode 27 or visionOS, pick an M-series Mac on either provider.

Migrate from AWS EC2 Mac to Figment

Capture the AWS side

• Export instance and host metadata: aws ec2 describe-instances --filters "Name=instance-type,Values=mac*" > mac-instances.json and aws ec2 describe-hosts > mac-hosts.json.
• Snapshot attached EBS volumes before cutover.
• Record Security Group rules, VPC/subnet placement and any VPC peering the Mac depends on.
• Note IAM roles attached via instance profile; plan a scoped replacement IAM user for the Figment host.
• Export signing certificates together with their private keys from Keychain Access as an encrypted .p12. Copy provisioning profiles as files from ~/Library/MobileDevice/Provisioning Profiles/; they are not Keychain items.
• List required secret names only. Restore values on the new host through your secret manager—do not dump the environment to a plaintext file.
• Save CI configuration: .github/workflows/, buildspec.yml, Buildkite pipelines, Fastlane lanes.

Bring up Figment

• Request the Figment M-series or Managed Build Mac configuration you need; the quote confirms chip, RAM, disk, term and the access method.
• Install your toolchain (Homebrew, the exact Xcode release, Fastlane, Ruby/Node versions pinned to your project).
• Import the .p12 into the login keychain; drop provisioning profiles into the profiles directory.
• Re-register self-hosted runners (GitHub Actions, Buildkite, GitLab) with new tokens.
• If the Mac must reach private AWS resources, build a site-to-site VPN or an SSM tunnel rather than exposing RDS/ElastiCache publicly.
• Update your own firewall, artifact store and CI service allowlists with the Figment egress address. Apple does not publish a TestFlight/App Store Connect allowlist to configure.

Cut over

• Run both hosts in parallel for at least three builds covering clean, incremental and release-signing paths.
• Validate a real TestFlight upload end-to-end from Figment.
• Release the AWS Dedicated Host only after its 24-hour minimum has elapsed.

Choose the Figment plan

If you want us to size a configuration or confirm current availability for a physical Mac, request a quote and we will respond with the exact chip, RAM, disk and term.

View Figment Mac plans → · Start a Figment Mac →

Related: Figment vs HostMyApple · Figment vs Mac Mini Vault · VM vs dedicated Mac hosting