RunsOn c8a.xlarge · EBS gp3 vs GitHub 4-core x64
RunsOn does the Rust and TypeScript builds for 8.1–12.8× less and 1.5–2.0× faster. Both 4 vCPU x64.
Price per minute 7.9× cheaper on RunsOn: $0.0015 vs $0.0120the builds are not all compared on cost yet
Rust build 12.8× cheaper on RunsOn: $0.0036 vs $0.0457RunsOn 2.0× faster: 1m 50s vs 3m 38s
TypeScript build 8.1× cheaper on RunsOn: $0.0032 vs $0.0258RunsOn 1.5× faster: 1m 12s vs 1m 44s
Docker build 12m 30s GitHub only · RunsOn not run
RunsOn c8a.xlarge · EBS gp3
jobs ran from Amazon.com (US)
GitHub 4-core
jobs ran from Microsoft (US)
Side by side
Every measurement both runners share, cost first. The dot marks the better value; gap is the larger over the smaller.
| Measurement | RunsOn | GitHub | Gap |
|---|---|---|---|
| Cost | |||
| Cost per Rust build | $0.0036 | $0.0457 | 12.8× RunsOn |
| Cost per TypeScript build | $0.0032 | $0.0258 | 8.1× RunsOn |
| Cost per Docker build | n/a: Not run: the PostHog Docker build needs more than 8 GB of memory; it ran out of memory on 8 GB runners. | $0.1746 | — |
| Price per minute | $0.0015 | $0.0120 | 7.9× RunsOn |
| Real builds, wall time | |||
| Rust build (spotify-player) | 1m 50s | 3m 38s | 2.0× RunsOn |
| TypeScript build (opencode) | 1m 12s | 1m 44s | 1.5× RunsOn |
| Docker build (PostHog) | n/a: Not run: the PostHog Docker build needs more than 8 GB of memory; it ran out of memory on 8 GB runners. | 12m 30s | — |
| CPU & memory | |||
| PassMark single-thread (points) | 3,785 | 2,895 | 1.3× RunsOn |
| PassMark CPU Mark, all cores (points) | 13,390 | 7,511 | 1.8× RunsOn |
| sysbench CPU, 1 thread (events/s) | 5,007 | 4,076 | 1.2× RunsOn |
| 7-Zip compression (MIPS) | 38,149 | 18,743 | 2.0× RunsOn |
| Memory read bandwidth | 36.3 GiB/s | 24.1 GiB/s | 1.5× RunsOn |
| Measurement | RunsOn | GitHub | Gap |
|---|---|---|---|
| Workspace disk (fio) | |||
| Random read, 4k QD32 (IOPS) | 2,994 | 19,575 | 6.5× GitHub |
| Random write, 4k QD32 (IOPS) | 2,991 | 11,216 | 3.7× GitHub |
| Random read, 4k QD1 (IOPS) | 1,773 | 7,286 | 4.1× GitHub |
| QD1 read latency, p99 | 897 µs | 241 µs | 3.7× GitHub |
| Sequential read | 401 MiB/s | 394 MiB/s | ≈ even |
| Sequential write | 401 MiB/s | 244 MiB/s | 1.6× RunsOn |
| Sync write, 4k + fdatasync (IOPS) | 540 | 4,163 | 7.7× GitHub |
| Network, cache & Docker | |||
| Download from GitHub | 169 MiB/s | 95 MiB/s | 1.8× RunsOn |
| Cache restore, 4 GB (actions/cache) | 18s | 41s | 2.3× RunsOn |
| Cache save, 4 GB (actions/cache) | 16s | 34s | 2.1× RunsOn |
| docker pull | 4.9s | 16.4s | 3.4× RunsOn |
| Database | |||
| pgbench read/write (TPS) | 3,096 | 4,414 | 1.4× GitHub |
| pgbench read-only (TPS) | 150,407 | 40,172 | 3.7× RunsOn |
| Queue | |||
| Queue under burst load | 16s | 3s | 5.3× GitHub |
- ≈ even: under 5% apart, or a build inside both runners' run-to-run spread
- Queue: identical jobs queued at the same instant (burst suite)
- 2 missing on one side or both
- RunsOn at EC2 spot, its default (7-day average of the cheapest us-east-1 zone's spot price, fetched 16 Nov 2026); its jobs ran on spot
- Cost: a managed SaaS runner against RunsOn self-hosted on your AWS account; the effort of running it was not measured