Data platform cloud cost estimate

Describe the data you keep, what your queries scan and the batch jobs you run, and see what the same setup costs each month on Azure, AWS, Google Cloud and Oracle Cloud, using each vendor's public list prices. A pay-per-use analytics lake: raw and curated files in object storage, serverless SQL billed by data scanned, scheduled batch transforms, and results sent out to users or other systems.

RegionCurrency

Start from an example

Step 1 of 3: Data

Step 01 / 03

Estimated monthly costIndia (Mumbai / Pune)

₹53,291a month

Google Cloud Lowest

Lowest of 3 at these inputs, 4.7% below the next.

Monthly total by provider

Google Cloud

₹53,291Lowest

Amazon Web Services

₹55,917

Microsoft Azure

₹86,248

Oracle Cloud

₹24,853Incomplete

Estimated monthly cost

Prices refreshed 30 Oct 2026 · Amounts are per month in INR.

Monthly cost by service line and provider, in INR
Service line	Google Cloud	Amazon Web Services	Microsoft Azure	Oracle Cloud
Object storage	22,437.99Google Cloud Storage (Standard, regional)	24,389.12Amazon S3 Standard	19,511.30Azure Blob Storage (Hot, LRS)	24,852.61OCI Object Storage (Standard)
Formula Google Cloud Google Cloud Storage (Standard, regional) Data stored10,240 GB-month × ₹2.19 = ₹22,437.9910 TB × 1024 GB a TB = 10,240 GB Request charges for files are not included; they are small next to storage for most apps. Vendor price page Amazon Web Services Amazon S3 Standard Data stored10,240 GB-month × ₹2.38 = ₹24,389.1210 TB × 1024 GB a TB = 10,240 GB Request charges for files are not included; they are small next to storage for most apps. Vendor price page Microsoft Azure Azure Blob Storage (Hot, LRS) Data stored10,240 GB-month × ₹1.91 = ₹19,511.3010 TB × 1024 GB a TB = 10,240 GB Request charges for files are not included; they are small next to storage for most apps. Vendor price page Oracle Cloud OCI Object Storage (Standard) Data stored10,240 GB-month × ₹0.0000 = ₹24,852.6110 TB × 1024 GB a TB = 10,240 GBVolume tiers applied, so the cost is below quantity × unit price. Request charges for files are not included; they are small next to storage for most apps. Vendor price page
SQL on the data lake	13,575.97BigQuery (on-demand, external tables on Cloud Storage)	9,527.00Amazon Athena	11,432.40Azure Synapse serverless SQL pool	No direct equivalent
Formula Google Cloud BigQuery (on-demand, external tables on Cloud Storage) Analysis (data scanned)20 TB × ₹0.0000 = ₹13,575.9720 TB scanned a monthVolume tiers applied, so the cost is below quantity × unit price. Vendor price page Amazon Web Services Amazon Athena Data scanned20 TB × ₹476.35 = ₹9,527.0020 TB scanned a month Vendor price page Microsoft Azure Azure Synapse serverless SQL pool Data processed20 TB × ₹571.62 = ₹11,432.4020 TB scanned a month Vendor price page Oracle Cloud No direct equivalent Oracle has no SQL service that bills per amount of data scanned. Autonomous AI Lakehouse can query object storage, but it is billed per ECPU-hour while the database runs, which is a different model from pay-per-scan, so the two are not compared.
Batch data transform	14,990.47Google Cloud Dataflow (batch)Workers of 4 vCPUs and 16 GB	20,959.40AWS Glue ETL jobsWorkers of 4 vCPUs and 16 GB	54,160.99Azure Data Factory mapping data flows (General Purpose)Workers of 4 vCPUs and 16 GB	Can't be sized
Formula Google Cloud Google Cloud Dataflow (batch) Batch worker vCPU2,000 vCPU-hour × ₹5.87 = ₹11,737.26500 worker-hours × 4 vCPUs Batch worker memory8,000 GB-hour × ₹0.4067 = ₹3,253.20500 worker-hours × 16 GB Workers of 4 vCPUs and 16 GB Vendor price page Amazon Web Services AWS Glue ETL jobs Glue DPU (4 vCPU, 16 GB)500 DPU-hour × ₹41.92 = ₹20,959.40500 worker-hours × 1 DPU Workers of 4 vCPUs and 16 GB Vendor price page Microsoft Azure Azure Data Factory mapping data flows (General Purpose) Data flow vCore (General Purpose)2,000 vCore-hour × ₹27.08 = ₹54,160.99500 worker-hours × 4 vCores Workers of 4 vCPUs and 16 GB Vendor price page Oracle Cloud OCI Data Integration Oracle bills a workspace by the hour plus data processed, not by compute workers, and does not say how many GB an hour of work handles, so it can't be sized from worker-hours. Vendor price page
Internet data transfer	2,286.48Google Cloud network egress (Premium Tier)	1,041.30AWS data transfer out	1,143.24Azure bandwidth (Microsoft network routing)	0.00OCI outbound data transfer
Formula Google Cloud Google Cloud network egress (Premium Tier) Data transfer out to India200 GB × ₹11.43 = ₹2,286.48200 GB a month Vendor price page Amazon Web Services AWS data transfer out Free data transfer out (all regions)100 GB × ₹0.0000 = ₹0.0000the first 100 GB a month are free Data transfer out to the internet100 GB × ₹10.41 = ₹1,041.30200 GB − 100 GB free Vendor price page Microsoft Azure Azure bandwidth (Microsoft network routing) Data transfer out to the internet200 GB × ₹0.0000 = ₹1,143.24200 GB a monthVolume tiers applied, so the cost is below quantity × unit price. Vendor price page Oracle Cloud OCI outbound data transfer Data transfer out (from Asia Pacific)200 GB × ₹0.0000 = ₹0.0000200 GB a month Vendor price page
Monthly total	53,290.91Lowest	55,916.82+4.9%	86,247.93+61.8%	24,852.61IncompleteTotal excludes: SQL on the data lake and Batch data transform

Assumptions

Region: India (Mumbai / Pune)
Currency: INR
Prices: On-demand list prices
Hours a month: 730
Not included: Support plans, taxes, negotiated discounts

What this workload means

Data stored: 10,240 GB
Data scanned a month: 20 TB
Transform worker-hours a month: 500 hours
Data out: 200 GB

How the quantities were worked out

Storage and queries are billed per month. Transform workers are billed only for the hours a job runs. On-demand list prices, no discounts or commitments, and no free tiers beyond what the vendor price list includes.
Storage is priced per GB and one terabyte is counted as 1,024 GB. Azure and AWS price queries per TB and Google per TiB; all three are treated as the same size.
One transform worker is 4 vCPUs and 16 GB of memory, the size of one AWS Glue DPU. Azure is priced by its 4 vCores (General Purpose) and Google by 4 vCPUs plus 16 GB of memory.
Files sit in standard object storage in one region. Request charges for storage are not included.
A query costs what it scans. Partitioning and column formats reduce the scan but are not modelled: enter the TB your queries actually read.
Oracle has no pay-per-scan SQL service and does not publish how much data an hour of Data Integration handles, so those two services are shown as gaps. Its estimate is incomplete and is never ranked cheapest.
Data catalog charges, orchestration, monitoring, support plans and taxes are not included.

Prices refreshed 30 Oct 2026

Assumptions	Monthly total
Google Cloud	53,290.91
Amazon Web Services	55,916.82
Microsoft Azure	86,247.93
Oracle Cloud	24,852.61