RunsOn c8g.xlarge · EBS gp3 vs Blacksmith 4 vCPU arm64
RunsOn does the Rust build for 4.7× less and 1.9× faster. Both 4 vCPU arm64.
Price per minute 3.8× cheaper on RunsOn: $0.0013 vs $0.0050the builds are not all compared on cost yet
Rust build 4.7× cheaper on RunsOn: $0.0057 vs $0.0267RunsOn 1.9× faster: 2m 43s vs 5m 04s
TypeScript build not yet RunsOn 1m 20s (1 run) · Blacksmith 2m 21s (5 runs)
Docker build 12m 12s Blacksmith only · RunsOn not run
RunsOn c8g.xlarge · EBS gp3
jobs ran from Amazon.com (US)
Blacksmith 4 vCPU
jobs ran from Secured Servers (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 | Blacksmith | Gap |
|---|---|---|---|
| Cost | |||
| Cost per Rust build | $0.0057 | $0.0267 | 4.7× RunsOn |
| Cost per TypeScript build | $0.0038 | $0.0132 | not yet |
| 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. | $0.0654 | — |
| Price per minute | $0.0013 | $0.0050 | 3.8× RunsOn |
| Real builds, wall time | |||
| Rust build (spotify-player) | 2m 43s | 5m 04s | 1.9× RunsOn |
| TypeScript build (opencode) | 1m 20s | 2m 21s | not yet |
| 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 12s | — |
| CPU & memory | |||
| PassMark single-thread (points) | 1,940 | 1,343 | 1.4× RunsOn |
| PassMark CPU Mark, all cores (points) | 3,905 | 3,778 | ≈ even |
| sysbench CPU, 1 thread (events/s) | 3,333 | 930 | 3.6× RunsOn |
| 7-Zip compression (MIPS) | 25,843 | 15,351 | 1.7× RunsOn |
| Memory read bandwidth | 19.3 GiB/s | 9.0 GiB/s | 2.1× RunsOn |
| Measurement | RunsOn | Blacksmith | Gap |
|---|---|---|---|
| Workspace disk (fio) | |||
| Random read, 4k QD32 (IOPS) | 2,992 | 155,975 | 52.1× Blacksmith |
| Random write, 4k QD32 (IOPS) | 2,994 | 24,132 | 8.1× Blacksmith |
| Random read, 4k QD1 (IOPS) | 1,703 | 24,897 | 14.6× Blacksmith |
| QD1 read latency, p99 | 889 µs | 39 µs | 22.7× Blacksmith |
| Sequential read | 401 MiB/s | 1.7 GiB/s | 4.4× Blacksmith |
| Sequential write | 401 MiB/s | 742 MiB/s | 1.9× Blacksmith |
| Sync write, 4k + fdatasync (IOPS) | 539 | 7,118 | 13.2× Blacksmith |
| Network, cache & Docker | |||
| Download from GitHub | 240 MiB/s | 85 MiB/s | 2.8× RunsOn |
| Cache restore, 4 GB (actions/cache) | 22s | 36s | 1.6× RunsOn |
| Cache save, 4 GB (actions/cache) | 14s | 21s | 1.5× RunsOn |
| docker pull | 7.3s | 13.7s | 1.9× RunsOn |
| Database | |||
| pgbench read/write (TPS) | 2,493 | 1,893 | 1.3× RunsOn |
| pgbench read-only (TPS) | 39,334 | 14,786 | 2.7× RunsOn |
| Queue | |||
| Queue under burst load | 18s | 15s | 1.2× Blacksmith |
- ≈ even: under 5% apart, or a build inside both runners' run-to-run spread
- not yet: one side has fewer than 3 runs, so the two values are shown but not compared
- Queue: identical jobs queued at the same instant (burst suite)
- 2 shown, not compared yet (under 3 runs on a side)
- 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 5 Oct 2026); the 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