Ray distributed computing framework
Ranked 2 of 3 distributed computing frameworks by recommenders
People recommend Ray for scaling Python workloads across machines, especially ML infrastructure, with built-in autoscaling. Critics say clusters add heavy overhead and cost, and that it is slow.
People recommend Ray for scaling Python workloads across machines, especially ML infrastructure, with built-in autoscaling. Critics say clusters add heavy overhead and cost, and that it is slow.
17 people gave an opinion
- recommend it
- 10
- criticize it
- 7
- approval share
- 59%
Recommended in 6 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: not recommended
- April 2026: recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 10 people recommend itNeeds 10 or more, met.
- 1.4 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 6 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Strong fit for ML infrastructure
“Ray is now becoming de-facto framework for compositional learning infrastructure”
Comment from 2 Dec 2025 in “Ray Keynote: de-facto for compositional learning and a distributed heterogeneous [video]” “have also noticed this pattern in Modal [1] and Ray [2] and all for it.”
Comment from 2 Feb 2026 in “Show HN: Stelvio – Ship Python to AWS” “there are libraries and clear documentation (e.g. Lightning, Ray, etc)”
Comment from 16 Dec 2025 in “In Defense of Matlab Code” - Comment from 17 Dec 2025 in “In Defense of Matlab Code”
Built-in autoscaling for clusters
“stitched together with Ray autoscaling nodes it can get really nice”
Comment from 13 Jul 2026 in “Ask HN: What Are You Working On? (July 2026)” “Ray has autoscaling, so it's very good for this kind of thing.”
Comment from 1 Nov 2025 on Hacker News
What people warn about
Each mark is one comment
Cluster overhead and high costs
“managing the infrastructure overhead of Celery or Ray is a massive time sink”
Comment from 14 Mar 2026 in “Show HN: I built Wool, a lightweight distributed Python runtime” “ray clusters don't scale well and end up costing you more money”
Comment from 30 Jan 2026 in “Quack-Cluster: A Serverless Distributed SQL Query Engine with DuckDB and Ray” Too slow compared with alternatives
“it's faster than prefect and ray so I can use it for more stuff without having to import a 10000000 line library”
Comment from 19 Mar 2026 in “What if Python was natively distributable?” “Dask and Ray were also way too slow”
Comment from 14 Nov 2025 in “650GB of Data (Delta Lake on S3). Polars vs. DuckDB vs. Daft vs. Spark”
Other distributed computing frameworks people recommend
The two most recommended other distributed computing frameworks, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Apache Spark | 18 | 25 | 42% | 9 | — | — |
| Ray distributed computing frameworkThis page | 10 | 7 | 59% | 6 | — | — |
| Apache Hadoop | 8 | 17 | 32% | 3 | — | — |
What people said
17 of the 19 comments that mention it, newest first. Each links to the original comment.
“stitched together with Ray autoscaling nodes it can get really nice”
“Personal finances with a TUI, love it.”
“So its essentially an AI trained on OpenSubtitles data.”
“it's faster than prefect and ray so I can use it for more stuff without having to import a 10000000 line library”
“spend weeks configuring things they didn't need”
“managing the infrastructure overhead of Celery or Ray is a massive time sink”
“Same cluster, zero interference, no runtime overhead.”
“have also noticed this pattern in Modal [1] and Ray [2] and all for it.”
“The Ray abstraction here is clever for bridging that gap”
“The Ray team even rejected Lambda support expressedly as far back as 2020”
“ray clusters don't scale well and end up costing you more money”
“so much more is solved by e.g. modern Python libs like Ray, Lighting”
“there are libraries and clear documentation (e.g. Lightning, Ray, etc)”
“One thing valuable is "Fault Tolerance"”
“Ray is now becoming de-facto framework for compositional learning infrastructure”
“Dask and Ray were also way too slow”
“Ray has autoscaling, so it's very good for this kind of thing.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.