←
Blacksmith821 posts
Blacksmith avatar

Blacksmith

@useblacksmith

The fastest way to run your GitHub Actions. Trusted by @supabase, @trychroma, @clerk and 10,000+ others.

◷ Joined January 2024▧ 175 Photos and videos
Blacksmith avatar
Blacksmith@useblacksmith·Oct 21

@sequoia is leading our Series C at a $1B valuation. Since January, we’ve gone from running 5M CI jobs a week to over 90M. Agents are writing an order of magnitude more code, and all of it still needs to be built and tested before it ships.

0:14
Blacksmith avatar
Blacksmith@useblacksmith·21h

Blacksmith is now in the a16z Speedrun Marketplace. Change one line of YAML and your existing GitHub Actions workflows run on faster bare-metal runners. More than 10,000 companies already run their CI this way.

Blacksmith avatar
Blacksmith@useblacksmith·Nov 13

Service containers now come up about 2x faster on Blacksmith. Every job with a services block waits on a container reporting healthy before its first step runs. This was dead time across millions of jobs a day.

Post attachment
Blacksmith avatar
Blacksmith@useblacksmith·Nov 9

Heading to KubeCon, November 9-12? Find Blacksmith at booth 272. Run GitHub Actions jobs up to 2x faster than GitHub-hosted runners. Change one line of YAML, keep your workflows, and leave autoscaling, caching, and infrastructure to us.

Post attachment
Blacksmith avatar
Blacksmith@useblacksmith·Nov 6

On November 5, 2026, our security team detected and contained unauthorized access to the system that stores the console output (job logs) of CI jobs run on Blacksmith runners. The access was revoked the same day, and a later attempt by the same party to regain access was blocked. Our investigation is ongoing. At this time there is no evidence that customer exposure goes beyond the job-log store. Based on what we know at this time, the incident impacted job-log data belonging to approximately 2.6% of customer accounts. We have contacted each of these customers directly with the specifics for their organization, and we continue to work with them. We do not believe any other Blacksmith customer needs to take action at this time. If our understanding changes, we will update this notice and notify any affected customers directly. We are continuing to actively investigate and to harden our systems, and we will publish a full report once that work is complete. Please reach out to security@blacksmith.sh if you have any further questions.

Blacksmith avatar
Blacksmith@useblacksmith·Nov 4

Blacksmith runs over 100 million jobs a week, so even a 70,000 vCPU burst is a small slice of our fleet. That scale is why we can offer unlimited concurrency, and why your hotfix never needs to wait in line. The math behind it: blacksmith.sh/blog/the-econo…

Blacksmith avatar
Blacksmith@useblacksmith·Nov 4

Two teams that each need 80 machines at peak can share about 110. Add thousands more teams and the bursts blend into a smooth, predictable load. That only works with a large fleet of your own and a growing set of customers sharing it, which is what we built.

Blacksmith avatar
Blacksmith@useblacksmith·Nov 4

The problem compounds through the day. Once pushes outpace the cap, the queue never clears and only gets deeper. Then prod breaks and the fix lands behind a thousand queued jobs. You either wait, or cancel everyone else's runs and ask the team to stop pushing until it merges.

0:18
Blacksmith avatar
Blacksmith@useblacksmith·Nov 4

The most obvious way to speed up CI is to run more of it at once. So teams shard: split a slow test suite into small jobs and run them in parallel. But every shard needs a slot, and under a cap most sit queued while the first few finish in minutes. Your CI is only as fast as its last shard.

0:16
Blacksmith avatar
Blacksmith@useblacksmith·Nov 4

You can't have fast CI with a concurrency cap. So Blacksmith doesn't have one, and we go to great lengths to keep it that way. Managed CI shouldn't mean managing limits. When your agents open 100 PRs at once, none of them should wait in a queue.

Blacksmith avatar
Blacksmith@useblacksmith·Nov 2

To restore a cache, Blacksmith runners pull from many storage servers at once, and under load those bursts kept stalling. We retuned the network path, and the median restore for caches over 256 MiB fell from 5.5 s to 2.8 s. The slowest 1% went from 35 s to 17 s.

Post attachment
Blacksmith avatar
Blacksmith@useblacksmith·Oct 30

30% less job time on some workloads, just from running in the right region. → We record what every job pulls: code, images, data → We measure how fast each region reaches it → We place each org nearest its traffic

0:16
Blacksmith avatar
Blacksmith@useblacksmith·Oct 27

Turborepo caching is now on by default for free on Blacksmith. 400+ teams are already using it. One monorepo that runs over a thousand Turborepo tasks a day now sees cached runs finish 97% faster.

0:09
Blacksmith avatar
Blacksmith@useblacksmith·Oct 23

Large downloads from GitHub are now 22% faster on Blacksmith runners (p90: 54 → 65 MB/s). We shipped a network optimization and measured it on production traffic, before and after. More coming soon.

Post attachment