Kernel vs TilionBrowser Benchmark Comparison

Kernel
Ahead in 0 of 2
vs
TilionPre-release
Ahead in 2 of 2

Tilion was ahead in both shared browser benchmarks.

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 for Browser lifecycle and higher is better for Browser throughput. Medians within 5% count as roughly even, and a benchmark gets no verdict when either provider completed fewer than 50% of runs.

BenchmarkKernelTilionResultRatio
Browser lifecycle
Total lifecycle time · Run October 30, 2026
0.22s
100% success
0.19s
100% success
Tilion faster
1.13x
Browser throughput
Actions per second · Run October 30, 2026
5.86 actions/s
99% success
6.91 actions/s
98% success
Tilion higher throughput
1.18x

Browser lifecycle

Run October 30, 2026

Time to create, connect to, navigate and release a headless browser session, over 100 iterations per run.

Total lifecycle time

Kernel

100% success
Median
0.22s
P95
0.26s
P99
0.27s
Kernel Browser lifecycle results →

Tilion

100% success
Median
0.19s
P95
0.24s
P99
0.27s
Tilion Browser lifecycle results →

Tilion had a 1.13x lower median total lifecycle time (0.19s vs 0.22s for Kernel).

Median total lifecycle time over time

15 runs from July 23, 2026 to October 30, 2026.

KernelTilion
View the full Browser lifecycle leaderboard →

Browser throughput

Run October 30, 2026

Sequential actions completed per second inside a running headless browser session: 10 actions across 100 sessions.

Actions per second

Kernel

99% success
Median
5.86 actions/s
P95
7.33 actions/s
P99
7.42 actions/s
Kernel Browser throughput results →

Tilion

98% success
Median
6.91 actions/s
P95
9.38 actions/s
P99
10.31 actions/s
Tilion Browser throughput results →

Tilion had a 1.18x higher median actions per second (6.91 actions/s vs 5.86 actions/s for Kernel).

Median actions per second over time

14 runs from July 23, 2026 to October 30, 2026.

KernelTilion
View the full Browser throughput 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.

Tilion was ahead in both shared browser benchmarks. In Browser lifecycle (October 30, 2026), Tilion had a median total lifecycle time of 0.19s versus 0.22s for Kernel, 1.13x faster. In Browser throughput (October 30, 2026), Tilion had a median actions per second of 6.91 actions/s versus 5.86 actions/s for Kernel, 1.18x higher throughput.

PartnersGoogle Cloud Run