Here are some examples that'll help you understand pricing of Firestore Enterprise edition in various scenarios.
Query pricing example
Scenario:A query filters on a field username
with the value ilovefirebase
in a collection containing 100 documents, where each document is
2KiB in size. Let's assume there is only one username
with the value ilovefirebase
. The same query is ran 1 million times.
| Activity | Standard edition (Auto-Indexed query) | Enterprise edition (Indexed query) | Enterprise edition (Unindexed Collection Scan) |
| Indexing Status | Uses automatically created indexfor the username lookup. | After manual index creationon the username field. | Does not use an index; scans the entire 200KiB user collection. |
| Read Units Per Query | Cost accrues 1 read. | Cost accrues 2 read units(1 for index scan and 1 for document read). | Cost accrues 50 read units(100 documents at 2KiB = 200KiB total scan. 200KiB / 4KiB per unit = 50 units). |
| Total Cost (per 1 Million Queries) | $0.30 per million queries(1 read @ $0.30/million read units). | $0.10 per million queries(2 read unit @ $0.05/million read units). | $2.50 per million queries(50 read units @ $0.05/million). |
Cost difference summary
- Indexed query:The same query on Enterprise edition costs $0.10 per million with a manually created index on the username field.
- Unindexed collection scan query:An unindexed collection scan query in Enterprise edition costs $2.50 per million, which is significantly higher than the $0.30 charged for the equivalent auto-indexed read in Standard edition. This illustrates that querying without indexes can lead to non-performant and costly execution.
- Cost advantage: $0.10 per million read unitsis three times cheaper than the Standard edition indexed read cost. This demonstrates the low base price of the Enterprise read unit when queries are optimized.
Real-time pricing example
For more information about pricing, see real-time pricing .
Scenario 1:One million clients query for the 10 most recent posts in a collection (indexed on a timestamp). Each post is 6KiB in size. Two of these posts are subsequently updated, resulting in real-time updates being pushed to the clients.
| Activity | Firestore Standard edition | Firestore Enterprise edition | Pricing Highlights |
| Pricing Model | Initial query and real-time updates are both charged per document read. | Initial query is charged using read unit (4KiB tranche). Real-time updates use a separate Real-time updates SKU. |
Enterprise leverages read unit pricing for cheaper initial reads |
| First Read (Initial sync of 10 posts with 6KiB documents each) | 10 readsare incurred per client (document size doesn't matter). | 21 read unitsare incurred per client (1 read unit for the index scan of 10 posts) (20 read units: 2 read units per 6KiB document * 10 documents) |
Enterprise consumes more units (21 versus 10) but at a much lower unit price. |
| First Read Cost (per 1 million ) | 10 reads * 1 million clients at $0.30/million: $ 3.00. | 21 read units * 1 million clients at 0.05/million: $ 1.05. | Initial read is 4 times cheaper in the Enterprise edition. |
| Real-time updates (2 updates to 6 KiB documents) | 2 readsare incurred (2 updated documents). | 4 real-time update unitsare incurred (2 real-time update units per 6KiB documents * 2 real-time updates). | Enterprise separates real-time updates into a dedicated SKU. |
| Real-time cost (per 1 million clients) | 2 reads * 1 million clients at $0.30/million: $ 0.60. | 4 real-time units * 1 million clients at $0.30/million: $ 1.20. | Real-time cost is higher in the Enterprise edition for this specific update scenario. |
| Total cost | $3.60($3.00 +$ $0.60). | $2.25($1.05 + $1.20). | Enterprise is cheaper ($2.25 vs $3.60)in this scenario involving large documents (6KiB) and high volume. |
Additional Real-time Listen queries pricing examples for Enterprise edition
For illustration, the costs in the following scenarios are calculated using the us-central1 rate of $0.05 per million read units and $0.30 per million real-time update units.
| Scenario (Clients, Document Size) | Phase 1: Initial Query (Sync) Details | Phase 1: Initial Query Consumption | Illustrative cost | Phase 2: Real-time Updates Details | Phase 2: Real-time Updates Consumption | Illustrative cost |
| 1 client, 2KiB docs | Query runs an unindexed collection scan of 10,000 sequential documents of 2KiB each. | 5,000 Read Units | $0.00025 (calculated based on $0.05/million) | Client receives 1,000 total documents using real-time updates, with a size of 2KiB per document. | 1,000 Real-time Update Units | $0.003 (Calculated based on $0.30/million) |
| 1,000 clients, 2KiB docs | 1,000 clients each run a query that does an unindexed collection scan of 10,000 sequential documents of 2KiB each. | 5 million Read Units | $0.25 (calculated based on $0.05/million) | 1,000 clients each receive 1,000 total documents using real-time updates, with a size of 2KiB per document. | 1 million Real-time Update Units | $0.30 (calculated based on $0.30/million) |
| 1,000 clients, 6KiB docs | 1,000 clients each run a query that does an unindexed collection scan of 10,000 sequential documents of 6KiB each. | 15 million Read Units | $0.75 (calculated based on $0.05/million) | 1,000 clients each receive 1,000 total documents using real-time updates, with a size of 6KiB each. | 2 million Real-time Update Units | $0.60 (Calculated based on $0.30/million) |
