Pydantic runs on RunsOn

Pydantic moved Logfire’s GitHub Actions workloads from a combination of GitHub-hosted runners, Depot, and Ubicloud to RunsOn-managed ephemeral EC2 runners in its AWS account on September 1, responding to CI costs that rose from $3,990 in June to $18,200 in August as AI-assisted development drove pull requests and workflow executions sharply higher. Although runner usage increased from 1.23 million minutes in August to 1.62 million in September, total monthly CI spending fell to $5,700, including RunsOn licensing and AWS compute, storage, caching, and transfer costs. RunsOn retains GitHub Actions as the workflow control plane while giving Pydantic greater control over runner types, networking, capacity, storage, and pricing, with most workloads using Graviton-based instances and x64 reserved for specific container builds and tests. The deployment combines local caching, persistent disks for BuildKit state, diversified EC2 capacity pools, and selective use of Spot and On-Demand instances, keeping interruption-sensitive merge, release, and deployment jobs on more reliable capacity. Pydantic concluded that effective cost control depends on using Spot capacity where retries are inexpensive, sizing runners for critical-path latency rather than simply choosing the lowest price, and placing caches close to compute while retaining them only when reuse is likely.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.