Buildkite

Personal plan renamed to Free Plan; Free and Pro plans now publish explicit usage caps: Free adds 5 users/10 concurrent jobs/2,000 VCPU-min/250k test executions; Pro capped at 50 users/250 agents/4,000 VCPU-min/1M test executions (previously unlimited).

"Model Providers" LLM proxy offering removed from pricing page — Anthropic Opus/Sonnet/Haiku input, output, cache-write, and cache-read token pricing no longer listed

Hosted Agent pricing moved to per-vCPU-minute billing (Linux $0.004/vCPU-min, Mac $0.02/vCPU-min), lowering effective rates (e.g. Linux Large $0.052->$0.032/min); non-profit/charity discount removed

Pro plan's self-hosted agent overage price rose from $2.50 to $3.50/agent/mo (+40%) new 'All Access Trial' plan added (full platform unlock, $0 for 30 days, no card)

Pro plan self-hosted agent overage rate increased from $2.50 to $3.50 USD/agent/mo (40% increase) for agents beyond the 10 included.

Pricing page heading changed from 'Everything you need for best-in-class software delivery' to 'Trial? More like Try All.' Personal plan tagline updated to 'Ship on your own' with bullet-point format.

Auto assignment feature removed from Test Engine plan comparison table across all plans. Several security and hosted agent features renamed: 'Git mirror caching' to 'Git mirror volumes'; 'Custom image support' to 'Custom agent images'; 'API token expiry policies' to 'Inactive API token revocation'; 'Session IP address pinning' to 'SSO session IP address pinning'; 'Software provenance' to 'Software (SLSA) provenance'.

Test Engine now includes Workflows feature with tiered limits across all plans [Personal: 1 workflow; Pro: 3 workflows; Enterprise: Volume discounts/Unlimited] API token expiry policies replaced with Inactive API token revocation, changing from policy-based expiry to automatic revocation of inactive tokens

Model Providers feature added - proxy for LLM model requests to Anthropic with pass-through pricing (Opus 4.5, Sonnet 4.5, Haiku 4.5) Personal plan introduced as a free tier ($0/mo) with limited features including 3 concurrent jobs, 50k test executions, 1GB storage, and 500 minutes of hosted agents

Mac M4 Medium RAM increased from 14GB to 28GB Mac M4 Large RAM increased from 28GB to 56GB

Test splitting feature added to Test Engine, available in both Pro and Enterprise plans API access overview feature democratized from Enterprise-only to both Pro and Enterprise plans

Free plan ($0) discontinued - no longer available as a standalone plan on pricing page Job minutes metric removed from Pro plan (was 40,000/month + 12,500 per user)

Enterprise plan restructured from fixed allocations (100,000 job minutes + 25,000 per user, 1,000,000 test executions) to volume-based pricing model Enterprise plan pricing changed from $35/user/month to Custom pricing