aws_timestreamwrite_table cost estimation
A serverless time-series table billed by writes, storage tier, and queries — no instances. 100M writes + 500 GB magnetic storage is ~$65/month.
An aws_timestreamwrite_table is a table in Amazon Timestream, a serverless time-series database. There's no cluster to provision; cost is usage across three meters: writes (~$0.50 per million 1 KB records), storage (memory store for recent hot data at a higher rate, magnetic store for older data at ~$0.03/GB-month), and queries (per GB scanned). 100M writes plus 500 GB of magnetic storage is ~$65/month.
The defining feature is the two-tier storage. Recent data lands in the in-memory store (fast queries, higher cost) and ages into the cheaper magnetic store per the table's retention settings. Tuning the memory-store retention — keeping only as much hot data as your queries actually need recent — is the main storage lever, since memory store costs far more per GB than magnetic.
Query cost scales with data scanned, so the same partitioning/scan-less discipline that controls Athena and BigQuery applies: filter by time range and dimensions so queries don't scan the whole table.
c3x prices the table from monthly write and storage volumes; queries are usage-driven.
Terraform example
A minimal but realistic configuration that C3X can estimate.
resource "aws_timestreamwrite_table" "metrics" {
database_name = aws_timestreamwrite_database.main.database_name
table_name = "device-metrics"
retention_properties {
memory_store_retention_period_in_hours = 24
magnetic_store_retention_period_in_days = 365
}
}Pricing dimensions
What you actually pay for when you provision aws_timestreamwrite_table.
| Dimension | Unit | What's being charged |
|---|---|---|
| Writes | per 1M records | Per million 1 KB write records ingested into the table. $0.50 per 1M writes → 100M = $50/month |
| Magnetic storage | per GB-month | Older data in the magnetic store. The memory store (recent hot data) costs more per GB. $0.03/GB-month → 500 GB = $15/month |
| Queries | per GB scanned | Query cost scales with data scanned. Usage-based — filter by time and dimensions. |
Sample C3X output
100M writes/month + 500 GB magnetic storage:
aws_timestreamwrite_table.metrics
├─ Writes (100M 1KB records) 100 × 1M-writes $50.00
└─ Magnetic storage 500 GB-month $15.00
Monthly $65.00Optimization tips
Common ways to reduce aws_timestreamwrite_table cost without changing the workload.
Tune memory-store retention
Large — memory store is the priciest tierThe in-memory store (recent hot data) costs far more per GB than magnetic. Keep only as much memory-store retention as your queries need for recent data; let everything else age into the cheaper magnetic store.
Filter queries by time and dimensions
Proportional to data not scannedQuery cost scales with data scanned. Always bound queries to a time range and filter dimensions so you scan only the relevant slice — the same discipline as Athena and BigQuery.
Batch writes
Per write reducedWrites bill per 1 KB record. Batching multiple measures into common records (multi-measure records) and writing in batches reduces the billed record count for the same data.
Set magnetic retention to compliance need
Per GB-month of stale data avoidedMagnetic storage is cheap but accumulates. Set magnetic_store_retention to what you actually query or must retain, not indefinitely.
FAQ
How is Amazon Timestream billed?
By usage, with no instances: writes (~$0.50 per million 1 KB records), storage (a pricey in-memory store for recent data plus cheap magnetic at ~$0.03/GB-month), and queries (per GB scanned). 100M writes + 500 GB magnetic is ~$65/month.
What's the biggest Timestream cost lever?
Memory-store retention. The in-memory store costs far more per GB than magnetic, so keeping only the recent data your queries need hot — and letting the rest age into magnetic — is the main storage saving. Time-bounded queries control query cost.
How does c3x estimate the cost?
From monthly write volume and storage in c3x-usage.yml, pricing writes and magnetic storage. Queries are usage-driven and modelled separately, since they depend on scan volume.
Related resources
Estimate this resource in your own Terraform
Free, open source, no API key. C3X parses your Terraform and shows line-item cost for every resource, including aws_timestreamwrite_table.