Comparison
Two ways to take data and AI all the way to production, on your side. This page compares graal and Cloudera criterion by criterion: what each one publishes, in its own words, dated and sourced.
At a glance
graal
With you: your Kubernetes or OpenShift, on-premises or with the hosting provider of your choice. No third-party service is needed to run it.
Cloudera
On AWS, Azure and Google Cloud, or on-premises, “in your data center or traditional data clusters”. Installing the Data Services with no Internet access is documented.
graal
Two open source products, under the Apache 2.0 license: graal Data Platform and graal Agentic Platform.
Cloudera
Software provided “under the terms of your agreement with Cloudera, Inc.”, on “a proven open-source foundation”.
graal
Annual vendor subscription per vCPU, unlimited users and agents, no usage-based billing.
Cloudera
In the cloud, an hourly rate per compute unit (CCU). On-premises, an annual subscription: “Contact Sales for pricing”.
graal
An MCP server drives the Data Platform, under a service account scoped to one project and with human approval. The Agentic Platform puts any agent to work on your processes.
Cloudera
Cloudera AI and Agent Studio, which consumes MCP servers. A read-only Apache Iceberg MCP server was announced in preview on 5 June 2025. Anywhere Cloud was announced on 19 August 2026.
Criterion by criterion
Every Cloudera cell links to its source and to the date we accessed it. Quotes stay in their original language.
| Criterion | graal | Cloudera |
|---|---|---|
| Where the platform runs | Your Kubernetes or OpenShift: on-premises, in a private cloud or with the hosting provider of your choice. Installed with Helm, without cluster administration rights; offline installation is possible. | In the cloud: “Cloudera on cloud services run on AWS, Azure, and Google Cloud.” On-premises: “in your data center or traditional data clusters”; the Data Services “Runs on ECS or dedicated OpenShift”. Anywhere Cloud, announced on 19 August 2026: “public clouds, sovereign clouds, private data centers, or isolated air-gapped networks”. |
| Who operates the control plane | You do. No SaaS and no telemetry: no third-party service is needed to run it. Operations by graal Services is a separate service that you choose. | In the cloud: “The Cloudera on cloud Control Plane and its corresponding infrastructure, security, and compliance are fully managed by Cloudera”, with workloads running in your cloud account. On-premises, the Data Services installation is launched from Cloudera Manager, including with no Internet access: “install Cloudera Data Services on premises in an air gap environment”. |
| Vendor and headquarters | GRAAL.SYSTEMS SAS, a French-law vendor, Paris. | Cloudera, Inc., in California: its press releases are datelined San Jose, and its license notice gives an address in Santa Clara. |
| Software license | Open source, Apache 2.0 license, for the Data Platform and the Agentic Platform alike. | “The Cloudera on prem (formerly the "Cloudera Data Platform - Private Cloud") Base offering is a collection of software components provided to you under the terms of your agreement with Cloudera, Inc.” Cloudera describes its approach as “leveraging a proven open-source foundation”. |
| Open formats | Parquet and Apache Iceberg on your S3 storage, standard SQL with Trino; low-code pipelines exported as Pandas or PySpark code. | Apache Iceberg: “Powered by the Apache Iceberg open table format”. On 4 June 2026, Cloudera announced its adoption of Apache Polaris, “an open source catalog built around the Apache Iceberg REST Catalog specification”. |
| Published pricing model | Annual vendor subscription per vCPU, unlimited users and agents, no usage-based billing. 8-week pilot; graal Services priced separately. | In the cloud: “Pricing is per Cloudera Compute Unit (CCU) which is a combination of Core and Memory.” Hourly rate, billed monthly or through pre-paid credits. On-premises: “Annual subscription with Business Support” and “Contact Sales for pricing”. |
| Generative AI and agents | MCP server built into the Data Platform, one service account per project, per-agent policies, human approval, revocation, audit. LLM gateway and RAG over your documents. Agentic Platform: workflows of agents and people, any agent, measured guarantees. | Cloudera offers AI Workbench and AI Inference: “Deploying and managing traditional models, large language models (LLMs), and generative AI models with high performance, security, and scalability”. In Agent Studio, “MCP servers act as tool families for specific services, such as Slack, Jira, databases, and web search”. An MCP server for Apache Iceberg through Impala, read-only, was “currently available in preview” on 5 June 2025; we found no more recent status. Anywhere Cloud, announced on 19 August 2026, is in limited availability according to the press, which reports general availability planned for 30 November 2026.
|
A cell older than 90 days is re-checked or removed. Has something changed? Write to contact@graal.systems and we will correct it.
When to choose graal
graal installs with Helm on your Kubernetes or OpenShift, without cluster administration rights, and brings together connectors, pipelines, orchestration and compute, an Iceberg lakehouse with Trino SQL, notebooks, ML and generative AI.
The vendor subscription is computed per vCPU, with unlimited users and agents and no usage-based billing: the bill varies neither with the number of queries nor with the number of people or agents.
Both platforms are open source, under the Apache 2.0 license. Low-code pipelines export as Pandas or PySpark code, and the data stays in Parquet and Iceberg on your storage.
A service account scoped to one project, per-agent policies, human approval for sensitive actions, immediate revocation and an audit log: your agents create, run and schedule without holding the keys to the platform.
When Cloudera suits you
Cloudera offers “traditional data clusters” on-premises, and compares its Base edition with CDH and HDP.
“Cloudera provides guaranteed long-term support, stability fixes, and enterprise maintenance for foundational existing Cloudera platform footprints through 2032 with version 7.3.2.”
“Cloudera on cloud hourly rates and Cloudera on premises annual subscriptions are aligned for hybrid cloud and multi-cloud flexibility.”
Anywhere Cloud targets “isolated air-gapped networks”, with natural-language search, agents and inference deployed “securely with zero external enterprise data exposure”.
Moving from Cloudera to graal
graal Services support the cut-over: inventory of your jobs, order of migration, team training.
Step 01
The HDFS connector reads your files where they sit, through WebHDFS or RPC: your data stays on the cluster during the move.
Step 02
Your PySpark jobs are dropped into a graal job. Spark runs on your Kubernetes, distributed when the size calls for it.
Step 03
You rewrite your tables as Iceberg on your S3 storage, then query them in SQL with Trino.
Step 04
Chains, cron and triggers are recreated in graal’s orchestration, with retries, backfills and alerts.
Frequently asked questions
Yes. The HDFS connector reads your files on the cluster, your PySpark jobs are dropped into graal and your tables are rewritten as Iceberg. graal Services support the cut-over, step by step.
Yes: offline installation loads the images from a local registry, on a network with no Internet access.
An annual vendor subscription per vCPU, with unlimited users and agents and no usage-based billing. The amount depends on your vCPU quota and the services you choose: you receive it after a scoping conversation.
Yes, through the Data Platform’s MCP server, under a service account scoped to one project. Sensitive actions wait for a manager’s approval, and every call is visible live and in the audit log.
The 8-week pilot puts graal to the test on your use case, in your infrastructure.