Head to head · Figures checked September 2026
Apache Gluten vs Databricks Photon
ETL Compare staff · Figures checked September 2026 · Sourced from vendor docs, project pages and public benchmarks · Published 29 September 2026
In brief
Apache Gluten has the higher ETL fit score on our published weights (3.9 against 3.4 out of 5). Apache Gluten scores higher on adoption effort, platform and instance portability, cost model transparency and published evidence; Databricks Photon scores higher on tuning and operating burden and maturity and community. They tie on stage coverage. Photon only runs inside Databricks, so for most teams the platform they use decides this comparison before the scores do.
On our weights Apache Gluten ranks 1st of 7 and Databricks Photon ranks 6th of 7. Which one fits depends on the criteria below and on the platform you run.
Apache Gluten (with Velox) is an open-source plugin that offloads Spark SQL execution to a native C++ engine (Velox or ClickHouse). Databricks Photon is a Databricks-native vectorized C++ query engine inside the Databricks Runtime. Neither requires changes to Spark SQL or DataFrame code, according to its own documentation.
3.9 / 5
ETL fit score (editorial assessment, 0-5)ETL fit score (editorial assessment, 0-5)
Rank 1 of 7
Highest ETL fit score: open-source native engine for self-managed Spark
3.4 / 5
ETL fit score (editorial assessment, 0-5)ETL fit score (editorial assessment, 0-5)
Rank 6 of 7
Best for teams already on Databricks
How do Apache Gluten and Databricks Photon score on each criterion?
| Criterion | Gluten + Velox | Photon | Higher score |
|---|---|---|---|
| Adoption effort 20% | 3.8 | 2.6 | Gluten + Velox |
| Stage coverage 20% | 3.8 | 3.8 | Tie |
| Tuning and operating burden 15% | 2.6 | 4.6 | Photon |
| Platform and instance portability 10% | 4.3 | 1.6 | Gluten + Velox |
| Cost model transparency 10% | 5.0 | 3.2 | Gluten + Velox |
| Published evidence 15% | 3.8 | 3.2 | Gluten + Velox |
| Maturity and community 10% | 4.4 | 4.6 | Photon |
| ETL fit score | 3.9 | 3.4 | Gluten + Velox |
Why each score
- Adoption effort
- Gluten + Velox (3.8): No application code changes; you add the Gluten JAR, set spark.plugins, enable off-heap memory and switch the shuffle manager, per the Velox getting-started page.
- Photon (2.6): One checkbox (or on by default on serverless) with no code changes if you already run on Databricks; for a team on EMR or self-managed Spark it means moving to Databricks first.
- Stage coverage
- Gluten + Velox (3.8): Offloads execution to the Velox native engine with a columnar shuffle manager; spilling is documented as experimental.
- Photon (3.8): Documents native scans (Parquet, Delta, CSV, JSON), filters, joins, aggregates, windows and writes, with fallback to the Spark runtime; no UDF, RDD or stateful streaming support.
- Tuning and operating burden
- Gluten + Velox (2.6): You size off-heap memory yourself (the docs example uses 20g) and rely on community channels; no vendor support contract is described on the project pages.
- Photon (4.6): Managed by Databricks inside its runtime, so there is no plugin for your team to install or version.
- Platform and instance portability
- Gluten + Velox (4.3): Supports Spark 3.4, 3.5, 4.0 and 4.1 on x86_64 and aarch64 Linux on standard CPU instances; managed-platform guides are not listed, so you install it yourself.
- Photon (1.6): Available only inside Databricks (on AWS, Azure and Google Cloud).
- Cost model transparency
- Gluten + Velox (5.0): Apache License 2.0, no license fee; cost is your existing compute plus engineering time.
- Photon (3.2): Databricks publishes DBU rates, but Photon compute consumes DBUs at different rates, so the net effect has to be modelled per job.
- Published evidence
- Gluten + Velox (3.8): Publishes TPC-H and TPC-DS results (Velox backend 2.71x overall, tested June 2023) with the benchmark named; results come from the project.
- Photon (3.2): Databricks documents where Photon does not help (queries under two seconds, UDFs); performance comparisons outside Databricks are not possible because it only runs there.
- Maturity and community
- Gluten + Velox (4.4): Apache top-level project since March 2026, started by Intel and Kyligence in 2022, with contributors including Alibaba Cloud, Meituan, Microsoft, IBM and Google.
- Photon (4.6): The default engine on Databricks serverless and SQL warehouses, backed by Databricks.
What do Apache Gluten and Databricks Photon cost, as published?
| Gluten + Velox | Photon | |
|---|---|---|
| License | Apache License 2.0 | Commercial (part of Databricks) |
| Published price | No license fee | Databricks DBU pricing; Photon changes DBU consumption |
| Instances | Standard CPU instances (x86_64 or aarch64) | Supported Databricks instance types, including Graviton |
| Runs on | Self-managed Spark 3.4 to 4.1 on Linux, on any platform where you control Spark config | Databricks only (AWS, Azure, Google Cloud) |
| Code changes | None; JAR plus Spark configuration | None on Databricks; platform move otherwise |
Prices and terms as published on the pages we reviewed, 27 September 2026. Neither vendor's figure is a quote.
Sources: gluten.apache.org, Velox backend getting started, apache/incubator-gluten on GitHub, Photon documentation · Fetched 27 Sep 2026
Where are they documented to run?
| EMR | Databricks | Google Cloud (Dataproc) | AWS Glue | Self-managed Spark / Kubernetes | |
|---|---|---|---|---|---|
| Gluten + Velox | Not documentedself-install | Not documented | Not documentedself-install | Not documented | Documented |
| Photon | Not available | Only platform | Not available | Not available | Not available |
Documented means the vendor or project lists the platform on the pages we reviewed. Self-install means you can usually add an open-source plugin to a platform that lets you set Spark configuration and classpath, but the project does not publish a guide for that platform. None of the vendor pages we reviewed list AWS Glue.
Which job stages does each address?
| Accelerator | Read | Transform | Shuffle | Spill | Write |
|---|---|---|---|---|---|
| Gluten + Velox | Partial | Documented | Documented | Partial | Not stated |
| Photon | Documented | Documented | Not stated | Not stated | Documented |
- Documented: the vendor's own documentation says it addresses this stage
- Partial: indirect or experimental, or covered only by an end-to-end claim
- Not stated: not found on the pages we reviewed
This shows what each vendor says, not what we measured. Sources are listed on each review.
Choose Apache Gluten if
- You want no license fee: Gluten is released under the Apache License 2.0
- You want the most widely contributed open-source native engine in this set, an Apache top-level project since March 2026
- You run on self-managed Spark or Kubernetes
Choose Databricks Photon if
- Your jobs already run on Databricks and you want no plugin to install or version
- You want the engine that is the default on Databricks serverless and SQL warehouses
- You run on Databricks
Frequently asked questions
Which has the higher ETL fit score, Apache Gluten or Databricks Photon?
Apache Gluten, with 3.9 against 3.4 out of 5 on our published weights. Apache Gluten scores higher on adoption effort, platform and instance portability, cost model transparency and published evidence and Databricks Photon on tuning and operating burden and maturity and community. They tie on stage coverage. The score measures fit for speeding up existing Spark ETL, not raw speed.
Do Apache Gluten and Databricks Photon run on the same platforms?
They share no documented platform on the pages we reviewed. Apache Gluten is also documented for self-managed Spark or Kubernetes. Databricks Photon is also documented for Databricks. Open-source plugins without a guide for a platform can often be self-installed where you control Spark configuration.
What do Apache Gluten and Databricks Photon cost?
Apache Gluten: No license fee (Apache License 2.0). Databricks Photon: Databricks DBU pricing; Photon changes DBU consumption (Commercial (part of Databricks)). For a like-for-like comparison, work out cost per run on one of your own jobs: see Spark cost per job, explained.