Daytona sandbox benchmark
Our hardware probe found Daytona runs each sandbox as a Docker container. We timed it through one SDK against 8 other sandbox providers: cold start, real agent work, everyday calls, pause and snapshots, bursts, machine speed, reliability and cost per task. Each result is a median with its p95 and 95% interval, on two tracks: the same 4 vCPU, 8 GB machine for every provider, and each provider's default machine.
Sandbox bench runs on Sandbox SDK, which the same team (opencoredev) builds. We may work with some of these providers. No provider paid for these results or decided how they were tested.
Same machine: 4 vCPU, 8 GB
- Cold start, median
- 1.3 s 8th of 9
- Tool replay, median
- 12 s 1st of 9
Default machines
- Cold start, median
- 1.5 s 8th of 9
- Tool replay, median
- 15 s 2nd of 9
Daytona on the same 4 vCPU, 8 GB machine
Every provider asked for Boat's machine, 4 vCPU and 8 GB, or its nearest size. Runs from 2026-10-22 to 2026-10-23, 9 providers. See every provider side by side on the full results.
- Asked for
- 4 vCPU, 8 GB, template daytona-large
- Measured inside
- 4 vCPU, 8 GB
- Cost per 1,000 tasks
- $1.36 (billed)
- Success rate
- 100.00% (976 of 976)
- Cost: $0.0014 per task, from its bill (billed over 10 tasks, low confidence). A task is: create a sandbox, run the 100-call tool replay, delete it. Cost per task comes from a separate cost run, billed per sandbox.
- List price: $0.331 per hour for this machine, compute only.
- Pause and resume: Daytona's container sandboxes have no pause: stop and start restarts the sandbox and ends every running process.
- Failures counted against Daytona: none.
Start-up
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Cold start (lower is better) | 1.3 sp95 2.1 s95%: 1.1 s to 1.3 s | 2.1 s | 1.1 s to 1.3 s | 8th of 9 · lowest median: Superserve 0.52 s |
Agent work
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Agent tool replay (lower is better) | 12 sp95 30 s95%: 12 s to 13 s | 30 s | 12 s to 13 s | 1st of 9 |
| npm install (lower is better) | 0.75 sp95 0.91 s95%: 0.73 s to 0.77 s | 0.91 s | 0.73 s to 0.77 s | 1st of 9 |
| pip install (lower is better) | 4.2 sp95 5.5 s95%: 4.0 s to 4.3 s | 5.5 s | 4.0 s to 4.3 s | 5th of 9 · lowest median: Tenki 3.3 s |
| git clone (lower is better) | 0.43 sp95 1.0 s95%: 0.42 s to 0.45 s | 1.0 s | 0.42 s to 0.45 s | 1st of 9 |
| Dev server ready (lower is better) | 2.6 sp95 6.6 s95%: 2.5 s to 2.6 s | 6.6 s | 2.5 s to 2.6 s | 5th of 9 · lowest median: Superserve 1.3 s |
| Preview URL first byte (lower is better) | 0.25 sp95 0.44 s95%: 0.25 s to 0.27 s | 0.44 s | 0.25 s to 0.27 s | 2nd of 9 · lowest median: Superserve 0.18 s |
| Edit, rebuild, reload (lower is better) | 0.31 sp95 0.76 s95%: 0.30 s to 0.34 s | 0.76 s | 0.30 s to 0.34 s | 2nd of 9, tied with Freestyle · lowest median: Freestyle 0.30 s |
Everyday calls
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Run a command (lower is better) | 0.13 sp95 0.30 s95%: 0.13 s to 0.14 s | 0.30 s | 0.13 s to 0.14 s | 2nd of 9, tied with E2B · lowest median: E2B 0.13 s |
| Write + read 1 KB (lower is better) | 0.18 sp95 0.52 s95%: 0.17 s to 0.18 s | 0.52 s | 0.17 s to 0.18 s | 2nd of 9 · lowest median: Superserve 0.10 s |
| Write + read 10 MB (lower is better) | 1.0 sp95 2.7 s95%: 1.0 s to 1.1 s | 2.7 s | 1.0 s to 1.1 s | 2nd of 9 · lowest median: Superserve 0.66 s |
| Reconnect (lower is better) | 0.30 sp95 0.71 s95%: 0.28 s to 0.32 s | 0.71 s | 0.28 s to 0.32 s | 1st of 9 |
| Destroy (lower is better) | 0.85 sp95 2.5 s95%: 0.80 s to 0.90 s | 2.5 s | 0.80 s to 0.90 s | 8th of 9 · lowest median: E2B 0.11 s |
Pause and snapshots
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Pause + resume (lower is better) | n/a | Daytona's container sandboxes have no pause: stop and start restarts the sandbox and ends every running process. | ||
| Take snapshot (lower is better) | 7.4 sp95 49 s95%: 6.8 s to 12 s | 49 s | 6.8 s to 12 s | 8th of 8 · lowest median: Superserve 0.39 s |
| Start from snapshot (lower is better) | 2.4 sp95 4.6 s95%: 2.3 s to 3.0 s | 4.6 s | 2.3 s to 3.0 s | 7th of 8 · lowest median: Superserve 0.55 s |
Bursts
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| 10 at once, all ready (lower is better) | 2.9 sp95 9.9 s95%: 2.3 s to 9.9 s | 9.9 s | 2.3 s to 9.9 s | 7th of 8, tied with boxd, Upstash Box · lowest median: Superserve 0.84 s |
| 25 at once, all ready (lower is better) | 5.3 sp95 5.6 s95%: 2.3 s to 5.6 s | 5.6 s | 2.3 s to 5.6 s | 7th of 9, tied with Boat, Upstash Box · lowest median: Superserve 0.79 s |
Machine speed
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| CPU, one core (higher is better) | 9.2M/sp95 10.1M/s95%: 9.1M/s to 9.3M/s | 10.1M/s | 9.1M/s to 9.3M/s | 9th of 9 · highest median: Tenki 19.4M/s |
| CPU, all cores (higher is better) | 25.9M/sp95 28.5M/s95%: 25.3M/s to 26.3M/s | 28.5M/s | 25.3M/s to 26.3M/s | 9th of 9 · highest median: Upstash Ephemeral Box 51M/s |
| Memory bandwidth (higher is better) | 29.9 GB/sp95 33.2 GB/s95%: 29.3 GB/s to 30.9 GB/s | 33.2 GB/s | 29.3 GB/s to 30.9 GB/s | 5th of 9 · highest median: Tenki 45.9 GB/s |
| Disk write (higher is better) | 1.4 GB/sp95 2.9 GB/s95%: 1.2 GB/s to 1.9 GB/s | 2.9 GB/s | 1.2 GB/s to 1.9 GB/s | 3rd of 9, tied with Upstash Box · highest median: Upstash Ephemeral Box 2.4 GB/s |
| Disk read (higher is better) | 17.3 GB/sp95 18.9 GB/s95%: 16.7 GB/s to 17.7 GB/s | 18.9 GB/s | 16.7 GB/s to 17.7 GB/s | 4th of 9 · highest median: Upstash Ephemeral Box 23.1 GB/s |
| 4K random write (higher is better) | 20.9K IOPSp95 30K IOPS95%: 19.6K IOPS to 22.1K IOPS | 30K IOPS | 19.6K IOPS to 22.1K IOPS | 2nd of 9 · highest median: Upstash Ephemeral Box 45.1K IOPS |
| 4K random read (higher is better) | 343K IOPSp95 403.6K IOPS95%: 337.3K IOPS to 364.1K IOPS | 403.6K IOPS | 337.3K IOPS to 364.1K IOPS | 7th of 9 · highest median: Boat 796.4K IOPS |
Quick picks on this track, all providers
| Need | Pick | Runner-up |
|---|---|---|
| Fastest cold start | Superserve 0.52 s | Upstash Ephemeral Box 0.68 s |
| Fastest agent work (tool replay) | Daytona 12 s | Superserve 13 s |
| Fastest 25-sandbox fan-out | Superserve 0.79 s | E2B 1.1 s |
| Fastest reconnect | Daytona 0.30 s | Superserve 0.33 s |
| Cheapest task | Upstash Ephemeral Box $0.00026 per task, $0.26 per 1,000 (billed) | Upstash Box $0.00034 per task, $0.34 per 1,000 (billed) |
| Most reliable | E2B, Tenki, Superserve, Upstash Box, Daytona, Upstash Ephemeral Box, Boat 100.00% (tied) | Freestyle 99.91% |
Daytona on its default machine
Each provider's default sandbox, so machine sizes differ. Runs from 2026-10-22 to 2026-10-23, 9 providers. See every provider side by side on the full results.
- Asked for
- 1 vCPU, 1 GB
- Measured inside
- 1 vCPU, 1 GB
- Cost per 1,000 tasks
- $0.64 (billed)
- Success rate
- 100.00% (973 of 973)
- Cost: $0.00064 per task, from its bill (billed over 10 tasks, low confidence). A task is: create a sandbox, run the 100-call tool replay, delete it. Cost per task comes from a separate cost run, billed per sandbox.
- List price: $0.067 per hour for this machine, compute only.
- Pause and resume: Daytona's container sandboxes have no pause: stop and start restarts the sandbox and ends every running process.
- Failures counted against Daytona: none.
Start-up
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Cold start (lower is better) | 1.5 sp95 2.8 s95%: 1.4 s to 1.6 s | 2.8 s | 1.4 s to 1.6 s | 8th of 9 · lowest median: E2B 0.50 s |
Agent work
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Agent tool replay (lower is better) | 15 sp95 34 s95%: 13 s to 18 s | 34 s | 13 s to 18 s | 2nd of 9 · lowest median: Superserve 13 s |
| npm install (lower is better) | 1.2 sp95 1.4 s95%: 1.2 s to 1.3 s | 1.4 s | 1.2 s to 1.3 s | 4th of 9 · lowest median: Boat 0.86 s |
| pip install (lower is better) | 4.6 sp95 6.2 s95%: 4.3 s to 4.7 s | 6.2 s | 4.3 s to 4.7 s | 7th of 9, tied with boxd · lowest median: Superserve 3.8 s |
| git clone (lower is better) | 0.62 sp95 0.97 s95%: 0.47 s to 0.87 s | 0.97 s | 0.47 s to 0.87 s | 2nd of 9, tied with Upstash Ephemeral Box · lowest median: Upstash Ephemeral Box 0.61 s |
| Dev server ready (lower is better) | 2.8 sp95 6.8 s95%: 2.6 s to 3.6 s | 6.8 s | 2.6 s to 3.6 s | 5th of 9 · lowest median: Superserve 1.8 s |
| Preview URL first byte (lower is better) | 0.28 sp95 0.46 s95%: 0.26 s to 0.31 s | 0.46 s | 0.26 s to 0.31 s | 3rd of 9 · lowest median: Superserve 0.19 s |
| Edit, rebuild, reload (lower is better) | 0.34 sp95 0.96 s95%: 0.33 s to 0.45 s | 0.96 s | 0.33 s to 0.45 s | 3rd of 9, tied with E2B · lowest median: Freestyle 0.31 s |
Everyday calls
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Run a command (lower is better) | 0.14 sp95 0.31 s95%: 0.13 s to 0.17 s | 0.31 s | 0.13 s to 0.17 s | 2nd of 9, tied with E2B · lowest median: E2B 0.14 s |
| Write + read 1 KB (lower is better) | 0.18 sp95 0.53 s95%: 0.17 s to 0.18 s | 0.53 s | 0.17 s to 0.18 s | 2nd of 9 · lowest median: Superserve 0.10 s |
| Write + read 10 MB (lower is better) | 1.2 sp95 2.5 s95%: 1.2 s to 1.3 s | 2.5 s | 1.2 s to 1.3 s | 2nd of 9 · lowest median: Superserve 0.64 s |
| Reconnect (lower is better) | 0.31 sp95 0.77 s95%: 0.29 s to 0.34 s | 0.77 s | 0.29 s to 0.34 s | 1st of 9 |
| Destroy (lower is better) | 0.96 sp95 3.2 s95%: 0.90 s to 1.1 s | 3.2 s | 0.90 s to 1.1 s | 8th of 9, tied with Upstash Box · lowest median: E2B 0.11 s |
Pause and snapshots
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| Pause + resume (lower is better) | n/a | Daytona's container sandboxes have no pause: stop and start restarts the sandbox and ends every running process. | ||
| Take snapshot (lower is better) | 10 sp95 336 s95%: 7.5 s to 15 s | 336 s | 7.5 s to 15 s | 8th of 8 · lowest median: E2B 0.34 s |
| Start from snapshot (lower is better) | 2.6 sp95 6.8 s95%: 2.3 s to 3.2 s | 6.8 s | 2.3 s to 3.2 s | 7th of 8 · lowest median: Superserve 0.56 s |
Bursts
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| 10 at once, all ready (lower is better) | 2.0 sp95 2.8 s95%: 1.7 s to 2.8 s | 2.8 s | 1.7 s to 2.8 s | 6th of 9, tied with Freestyle, boxd · lowest median: Superserve 0.65 s |
| 25 at once, all ready (lower is better) | n/a | In one of its three bursts of 25, one create timed out on Daytona's side (24 of 25 ready), so only 2 of 3 bursts finished and the size is not published; coming in the next update |
Machine speed
| Test | Median | p95 | 95% interval | Rank |
|---|---|---|---|---|
| CPU, one core (higher is better) | 9.1M/sp95 13.7M/s95%: 8.9M/s to 9.3M/s | 13.7M/s | 8.9M/s to 9.3M/s | 9th of 9 · highest median: Boat 18.8M/s |
| CPU, all cores (higher is better) | 6.9M/sp95 10.3M/s95%: 6.8M/s to 7M/s | 10.3M/s | 6.8M/s to 7M/s | 9th of 9 · highest median: Boat 51.7M/s |
| Memory bandwidth (higher is better) | 31.0 GB/sp95 57.0 GB/s95%: 30.3 GB/s to 31.9 GB/s | 57.0 GB/s | 30.3 GB/s to 31.9 GB/s | 5th of 9 · highest median: Upstash Ephemeral Box 41.1 GB/s |
| Disk write (higher is better) | 1.2 GB/sp95 2.9 GB/s95%: 1.1 GB/s to 1.3 GB/s | 2.9 GB/s | 1.1 GB/s to 1.3 GB/s | 3rd of 9 · highest median: Upstash Ephemeral Box 2.4 GB/s |
| Disk read (higher is better) | 16.2 GB/sp95 28.8 GB/s95%: 15.9 GB/s to 16.6 GB/s | 28.8 GB/s | 15.9 GB/s to 16.6 GB/s | 4th of 9 · highest median: Upstash Ephemeral Box 23.0 GB/s |
| 4K random write (higher is better) | 20.2K IOPSp95 29.6K IOPS95%: 18.6K IOPS to 20.9K IOPS | 29.6K IOPS | 18.6K IOPS to 20.9K IOPS | 2nd of 9 · highest median: Upstash Ephemeral Box 43.3K IOPS |
| 4K random read (higher is better) | 342.4K IOPSp95 774.7K IOPS95%: 323.9K IOPS to 348.6K IOPS | 774.7K IOPS | 323.9K IOPS to 348.6K IOPS | 7th of 9 · highest median: Boat 772.6K IOPS |
Quick picks on this track, all providers
| Need | Pick | Runner-up |
|---|---|---|
| Fastest cold start | E2B 0.50 s | Superserve 0.53 s |
| Fastest agent work (tool replay) | Superserve 13 s | Daytona 15 s |
| Fastest 25-sandbox fan-out | Superserve 0.75 s | E2B 0.87 s |
| Fastest reconnect | Daytona 0.31 s | Superserve 0.32 s |
| Cheapest task | Upstash Ephemeral Box $0.00013 per task, $0.13 per 1,000 (billed) | Upstash Box $0.00017 per task, $0.17 per 1,000 (billed) |
| Most reliable | E2B, Tenki, Superserve, Upstash Box, Daytona, Boat 100.00% (tied) | boxd 99.81% |
How to read these numbers
- Daytona was re-collected on Oct 22 after its account quota was raised from 10 vCPU / 10 GiB; earlier Daytona numbers were limited by that quota.
- Daytona's billed cost can still change for up to 48 hours.
- Every call starts from one server in Florida, US. Providers hosted farther away pay that round trip on every call; Boat (France) and boxd (Germany) pay a transatlantic one.
- There are two tracks. Same machine: 4 vCPU, 8 GB: Every provider asked for Boat's machine, 4 vCPU and 8 GB, or its nearest size. A provider marked “machine differs” measured something else inside. Default machines: Each provider's default sandbox, so machine sizes differ.
- Cost per task has a kind. Billed comes from the provider's bill, metered from its usage records at published rates, and estimate from our measured time at list price. Kinds are not directly comparable. Prices cover compute only at list rates: disk, transfer, subscriptions and credits are excluded.
- Pause and resume times come from the previous SDK run, because the newest runs skipped them. A burst size is published only when all three bursts finished.
- Failures caused by our own code are set aside, not counted. Every provider-side failure is kept.
- Ranks compare medians. Medians inside each other's 95% intervals count as tied, so a close rank can be noise.
More results
- E2B vs Daytona, side by side
- Upstash Box vs Daytona, side by side
- E2B sandbox benchmark
- Superserve sandbox benchmark
- Upstash Box sandbox benchmark
- Upstash Ephemeral Box sandbox benchmark
- Boat sandbox benchmark
- Freestyle sandbox benchmark
- Tenki sandbox benchmark
- Every sandbox provider we benchmark
- Every provider on the same 4 vCPU, 8 GB machine
- Every provider on its default machine
- Raw data: results.json · results.md · llms.txt