Mosaic vs TenkiSandbox Benchmark Comparison

Mosaic
Faster in 1 of 2
vs
Tenki
Faster in 0 of 2

Mosaic was faster in 1 of 2 shared sandbox benchmarks, and 1 was roughly even.

Shared benchmarks
2
Last run
October 30, 2026
Want the full run details?Sign up free →

Scorecard

Median of each benchmark's primary metric; lower is better. Medians within 5% count as roughly even, and a benchmark gets no speed verdict when either provider completed fewer than 50% of runs.

BenchmarkMosaicTenkiResultRatio
Burst TTI
Time to interactive · Run October 30, 2026
889ms
100% success
2.10s
100% success
Mosaic faster
2.37x
DAX
Total workload time · Run October 30, 2026
58.64s
100% success
60.78s
100% success
Roughly even
1.04x

Burst TTI

Run October 30, 2026

Time from sandbox creation to the first successful command, with 100 sandboxes launched at the same time.

Time to interactive

Mosaic

100% success
Median
889ms
P95
3.47s
P99
3.53s
Mosaic Burst TTI results →

Tenki

100% success
Median
2.10s
P95
2.18s
P99
2.18s
Tenki Burst TTI results →

Mosaic had a 2.37x lower median time to interactive (889ms vs 2.10s for Tenki).

Median time to interactive over time

15 runs from August 14, 2026 to October 30, 2026.

MosaicTenki
View the full Burst TTI leaderboard →

DAX

Run October 30, 2026

A real developer workload: install system packages, download Bun, clone the opencode repo, then run bun install and bun typecheck.

Total workload time

Mosaic

100% success
Median
58.64s
P95
58.64s
P99
58.64s
Mosaic DAX results →

Tenki

100% success
Median
60.78s
P95
60.78s
P99
60.78s
Tenki DAX results →

The medians are within 5% of each other (58.64s vs 60.78s), so this counts as roughly even.

Median total workload time over time

13 runs from August 14, 2026 to October 30, 2026.

MosaicTenki
View the full DAX leaderboard →

Methodology

  • One workload, every provider. The same script, payload, and model hit every provider, addressed the way its own API expects. Nobody gets a tuned or special path.
  • Repeated on a schedule. Every suite runs many iterations per provider, automated on a recurring schedule, so results reflect steady state rather than one lucky request.
  • Raw results published. Every run is committed as JSON to the public benchmarks repo, next to the code that produced it. Nothing is cherry-picked or held back.
  • How this comparison is built. Each benchmark compares the two providers' medians; results within 5% count as roughly even. A comparison is published only when both providers completed at least 50% of runs in 2 or more shared benchmarks run in the last 30 days.

Frequently asked questions

Have another question? Email us.

Mosaic was faster in 1 of 2 shared sandbox benchmarks, and 1 was roughly even. In Burst TTI (October 30, 2026), Mosaic had a median time to interactive of 889ms versus 2.10s for Tenki, 2.37x faster. In DAX (October 30, 2026), Mosaic had a median total workload time of 58.64s and Tenki 60.78s, which counts as roughly even.

PartnersGoogle Cloud Run