evaluate
The engine under every knitted R document reached 1.0 by making evaluation behave like the console.
A side-by-side editorial comparison of rush and Windmill — release velocity, themes, recent moves, and the top alternatives to consider.
Rush spent 2026 making distributed task ownership correct, starting with a breaking 1.0.
rush is the Redis-backed distributed computing backend behind mlr3's parallel tuning: workers pull tasks, push results and are tracked through heartbeats. Since 1.0.0 in March 2026 it has shipped five releases in four months, almost all correcting how tasks and workers are owned and identified under failure. The most recent lets each worker start in its own mirai call so it can spawn local daemons.
Windmill is dismantling the paywall between itself and the data stack
Three of the last six releases are directional: dbt projects running unmodified as a first-class runtime, BigQuery and Snowflake dropping out of the Enterprise tier, and AI sessions enabled by default in beta. Around them sits careful platform work — workspace lineage as the single deployment target, arbitrary-target Compare & Deploy, schema contracts for DuckLake consumers, and in-app git-to-Windmill sync. Entries are written with migration notes and upgrade paths, which is the tone of a product being run by people who expect it in production.
rush is the Redis-backed distributed computing backend behind mlr3's parallel tuning: workers pull tasks, push results and are tracked through heartbeats. Since 1.0.0 in March 2026 it has shipped five releases in four months, almost all correcting how tasks and workers are owned and identified under failure. The most recent lets each worker start in its own mirai call so it can spawn local daemons.
The through-line is moving responsibility from the controller to the worker. 1.0.0 rewrote the task lifecycle and dropped the seed mechanism; 1.2.0 moved fail_tasks(), finish_tasks(), pop_task() and push_running_tasks() from Rush onto RushWorker so only the worker that ran a task can mark it done, and gave worker ids random suffixes so they cannot collide. 1.1.0 had already removed phantom failed tasks when a worker crashes between evaluations.
With ownership and identity settled, the remaining pressure points are heartbeat and lost-worker detection; expect the next releases to tighten failure recovery rather than add API surface.
Three of the last six releases are directional: dbt projects running unmodified as a first-class runtime, BigQuery and Snowflake dropping out of the Enterprise tier, and AI sessions enabled by default in beta. Around them sits careful platform work — workspace lineage as the single deployment target, arbitrary-target Compare & Deploy, schema contracts for DuckLake consumers, and in-app git-to-Windmill sync. Entries are written with migration notes and upgrade paths, which is the tone of a product being run by people who expect it in production.
The company is aiming squarely at data teams and removing every obstacle between them and a self-hosted trial: bring the dbt project as-is, query the warehouse on the community edition, and let the AI build scripts and flows behind the existing draft-and-diff review gate. The governance work advances at the same pace, so the free surface widens without the deployment story loosening. Enterprise is being narrowed to Oracle, MS SQL, and scale rather than to the workloads data teams actually run.
With dbt lineage and DuckLake schema contracts both landed, the likely next step is unifying them — a single asset graph spanning dbt models and materialized tables, with contract validation running across the boundary.
Other Infra & APIs products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either rush or Windmill.
The engine under every knitted R document reached 1.0 by making evaluation behave like the console.
carrier's 0.3.0 quietly rewrites how crated functions see each other — and breaks code that relied on the old rules.
watcher is a thin filesystem-events binding whose releases are mostly build work.
gh spent its last two releases making failures and interruptions recoverable.
checkhelper grew from a check wrapper into a CRAN pre-submission auditor.
dockerfiler's 1.0.0 changes what its generated Dockerfiles actually contain.
See all rush alternatives → · See all Windmill alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Windmill is currently shipping more aggressively (velocity 8.8 vs 2.5), with 3 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. Windmill is currently shipping more aggressively (velocity 8.8 vs 2.5), with 3 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Infra & APIs products to evaluate alongside.
Top rush alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "rush alternatives" section above for the current picks, or visit /alternatives/rush for the full list with editorial commentary on each.
Top Windmill alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Windmill alternatives" section above for the current picks, or visit /alternatives/windmill for the full list with editorial commentary on each.