Fly Machines vs Unikraft Cloud
Choose Fly Machines for teams that want raw, per-second-billed Firecracker VMs they orchestrate themselves rather than a sandbox SDK; choose Unikraft Cloud for agent fleets that need thousands of sandboxes per server, millisecond stateful wake-ups and live forking of running environments. Fly Machines uses Firecracker microVM isolation; Unikraft Cloud uses microVM (modified Firecracker) isolation.
Key differences between Fly Machines and Unikraft Cloud
- Isolation: Fly Machines uses Firecracker microVM; Unikraft Cloud uses microVM (modified Firecracker).
- Self-hosting: Unikraft Cloud documents a self-hosted option; Fly Machines does not.
- Your own cloud account: Unikraft Cloud documents a bring-your-own-cloud deployment; Fly Machines does not.
- Company headquarters: Fly.io is based in Chicago, United States; Unikraft is based in Heidelberg, Germany.
- SDKs and interfaces: Only Unikraft Cloud lists JavaScript/TypeScript and Python.
What they have in common
- Both are hosted platforms.
- Both use microVM isolation.
- Both support snapshots.
- Both can pause and resume a sandbox.
- Both list Go, CLI and REST API.
Fly Machines vs Unikraft Cloud at a glance
| Attribute | Fly Machines | Unikraft Cloud |
|---|---|---|
| What it is | Fly.io's low-level API for Firecracker VMs that start and stop in under a second and can be suspended to disk; the primitive underneath Fly.io's Sprites product. | Millisecond-start microVMs built from Dockerfiles on a modified Firecracker VMM, with stateful scale-to-zero, live forking ("branching") and checkpoints, plus a Sandbox SDK for agents. |
| Type | Hosted platform | Hosted platform |
| Isolation | Firecracker microVM per Machine ("Compute jobs at Fly.io are virtualized using Firecracker"); runs on Fly.io's own hardware in 17 regions | Hardware-isolated microVM per instance with its own kernel, run on a "Modified Firecracker VMM" (KVM); sandbox egress passes through a separate "network shield" microVM that holds credentials |
| Open source | No | Not documented |
| Self-host / BYOC | No / No | Yes / Yes |
| SDKs | Go, CLI, REST API | JavaScript/TypeScript, Python, Go, CLI, REST API |
| Cold start (vendor claim) | Docs: starting a created Machine "usually takes well under a second"; creating one (image fetch, rootfs build) "maybe low double digit seconds"; resuming a suspended Machine "takes a few hundred milliseconds"; marketing claims "sub-second starts and stops" | Vendor claims "<10ms" cold start and scale-to-zero for all workloads, "branch & restore in < 10 ms", and "Fork a live virtual computer in 10 milliseconds"; the how-it-works page says apps are pre-initialised and snapshotted so a request to a dead-cold instance is answered "within a single round trip" |
| Max session | Not documented | Not documented |
| Persistence | Stopped Machines keep their rootfs assembled on the host for fast restart (ephemeral storage is a clean slate on each start); suspended Machines "preserve memory and disk state" and resume faster than stopped; attach a Fly Volume for persistent storage; volume snapshots | Stateful scale-to-zero (standby instances resume with memory intact); templates are snapshotted instances that preserve memory, open files and caches; branching clones a running sandbox including memory and volume state; checkpoints keep a history you can rewind to; block volumes; live migration between servers; "autokill" terminates long-running sandboxes |
| Snapshots / pause | Yes / Yes | Yes / Yes |
| GPU | GPU Machines have been offered; the GPU docs URL now redirects to a blog post that says Fly is "scaling back our GPU ambitions" but "we're not getting rid of them"; no GPU rates appear on the pricing page | Preview for enterprise customers only: requires the full (QEMU) instance type, at most one GPU per instance and you cannot choose which (example model NVIDIA Tesla T4); full VMs do not support branching or checkpointing |
| Networking | Private IPv6 network per organisation (WireGuard access); public IPv4/IPv6 and Fly Proxy routing on request; static egress IPs available (fly ips allocate-egress); inbound services optional | Each instance gets a private IP and .internal FQDN; service groups expose a public domain <name>.<metro>.unikraft.app; sandbox traffic is filtered by the network-shield microVM so "nothing reaches the internet directly" |
| Pricing | Billed by the second while running; stopped Machines pay only rootfs at $0.15/GB-month. Monthly equivalents in the cheapest regions: shared-cpu-1x 256 MB $2.19, 512 MB $3.69, 1 GB $6.70; shared-cpu-2x 2 GB $13.39; performance-1x 2 GB $33.00; performance-2x 4 GB $66.00; extra RAM $6/GB-month; volumes $0.15/GB-month; snapshots $0.08/GB-month after 10 GB free; 40% off reserved compute blocks; prices vary by region | Hobby $0/month (2 running and 5 standby instances), Team $39/month (8 running, 50 standby), Pro $199/month (24 running, 5,000 standby); Enterprise (dedicated hosts, on-prem, BYOC) custom |
| Free tier | $1 trial credit for 7 days (no card required); personal organisations with monthly bills under $5 are waived; no ongoing free tier | Hobby plan at $0/month ("Free forever"): 2 running instances, 5 standby instances, 2 vCPUs and 4 GiB memory in total |
| Compliance | SOC 2 Type 2; BAA and SOC 2 report available in the Compliance Package via the dashboard Trust Center; DPAs on request | SOC 2 Type II; HIPAA-compliant; independent third-party penetration tested (stated on unikraft.com; the trust centre at trust.unikraft.com requires an access request) |
| Regions | 17 regions: Amsterdam, Stockholm, Paris, Dallas, Secaucus, Frankfurt, São Paulo, Ashburn, Johannesburg, Los Angeles, London, Tokyo, Chicago, Singapore, San Jose, Sydney, Toronto | Default metros: Dallas, Frankfurt, San Francisco, Singapore, Washington DC; around two dozen further metros available on request |
| Company headquarters | Chicago, Illinois, United States | Heidelberg, Baden-Württemberg, Germany |
| Good fit for | Teams that want raw, per-second-billed Firecracker VMs they orchestrate themselves rather than a sandbox SDK | Agent fleets that need thousands of sandboxes per server, millisecond stateful wake-ups and live forking of running environments |
Verified November 14, 2026 from each provider’s documentation. Cold-start figures are vendor claims, not our measurements.
Which should you choose?
Reasons to choose Fly Machines
- You want per-second-billed Firecracker VMs with full control of lifecycle, region placement and networking
- You need suspend-to-disk and sub-second restarts for bursty or idle-heavy workloads
- You are already on Fly.io and want sandboxes on the same private network as your app
- You run long-lived or persistent workloads (volumes, no documented session cap)
Reasons to choose Unikraft Cloud
- You need very high sandbox density and millisecond resume from idle
- You want to fork a running agent environment (memory included) for parallel sub-agents
- You need a credential-holding egress proxy so sandboxed code never sees secrets
- You run on your own hardware or cloud and want the same stack on-prem or BYOC
Trade-offs to check first
- Fly Machines: No dedicated sandbox SDK: file-transfer and snapshot-fork APIs for agents are not documented, and command execution is offered through flyctl (
fly machine exec) - Fly Machines: Only an official Go client; other languages must be generated from the OpenAPI spec
- Unikraft Cloud: Self-serve plans cap running instances (2 on Hobby, 8 on Team, 24 on Pro); higher limits need the custom-priced Enterprise plan
- Unikraft Cloud: GPU support is an enterprise preview on a different (QEMU) instance type without branching or checkpoints
Both are worth piloting with your own image and region. See the full Fly Machines profile and Unikraft Cloud profile, or the complete comparison.
More comparisons
Frequently asked questions
Is Fly Machines or Unikraft Cloud better for AI agents?
What is the difference between Fly Machines and Unikraft Cloud?
Which is open source, Fly Machines or Unikraft Cloud?
How do Fly Machines and Unikraft Cloud isolate code?
How does Fly Machines pricing compare with Unikraft Cloud?
Do Fly Machines and Unikraft Cloud support GPUs?
full (QEMU) instance type, at most one GPU per instance and you cannot choose which (example model NVIDIA Tesla T4); full VMs do not support branching or checkpointing.Can you self-host Fly Machines or Unikraft Cloud?
Do Fly Machines and Unikraft Cloud support snapshots or pause and resume?
Sources
Check before you rely on it. Products change quickly, and this page reflects public documentation as of November 14, 2026. Confirm anything that matters in the provider’s own documentation.
Cite this page
Agent Sandboxes. (2026). Fly Machines vs Unikraft Cloud. agentsandboxes.org. Last reviewed November 14, 2026. https://agentsandboxes.org/compare/fly-machines-vs-unikraft-cloud/