# MongoDB Atlas

> mongodb.com | Databases and KV, Compute, Storage; 4 records: free tier, limits, identity cost and evidence.

HTML https://fshy.ajam.dev/p/mongodb/ | JSON https://fshy.ajam.dev/api/v1/entity/mongodb.json | how to read a record: https://fshy.ajam.dev/llms-full.txt

## MongoDB Atlas (free cluster, formerly M0)

- `databases-kv/mongodb-atlas-m0` | qualifies-with-caveat | evidence documented | binding per-ip-shaping | category document database
- needs card no, phone no, ID no, institution no | verified 2026-10-22T03:05:00.000Z (live) | stale on 2026-11-22
- url https://www.mongodb.com/docs/atlas/reference/free-shared-limitations/
- free tier: 0.5 GB of data storage on a shared cluster with three-node replication, one free cluster per project, free forever | unit: GB | metered on: stored-bytes | read on: 2026-10-22 | source: [1]
- per-object cap: 16 mebibytes per BSON document, which is the MongoDB engine limit and not a tier limit | bytes: 16777216 | source: [2]
- retention: paused automatically after 30 days with zero connections; resumable unless the cluster ran a version Atlas can no longer restore to | days: 30 | source: [1]
- egress: 10 GB in and 10 GB out per rolling seven-day period; exceeding it throttles the cluster and adds a one-second cooldown per connection | metered: yes | free allowance: 10 GB out per 7 days | source: [1]
- identity: email | detail: an Atlas account; no card on Free
- interface: rest, cli
- limits:
  - per-ip-shaping (binding): 100 operations per second, enforced by throttling with a one-second cooldown [1]
  - per-object-cap: 0.5 GB of data storage; storage size is not configurable on a free cluster [1]
  - request-rate: 10 GB in and 10 GB out per rolling 7-day period [1]
  - retention-clamped: paused after 30 days with zero connections [1]
- limits note: 500 connections maximum; replication factor 3; no sharding, no network peering, and only a subset of regions
- interface detail: the MongoDB wire protocol; a Data API for HTTPS access; the Atlas Administration API and the atlas CLI
- tooling: mongosh, any MongoDB driver, the atlas CLI
- composes with: a free document store that survives inactivity for a month, which is the longest idle clock in this class short of Aiven
- reliability: the page is explicit that Atlas may deactivate idle free clusters under its terms, and recommends exporting a copy before a long pause
- notes: 0.5 GB is a hard wall rather than an overage, and 100 ops/s is a throughput cap that no storage allowance can exceed. The 30-day idle clock is the most forgiving in this class and it is still a clock.
- evidence note: both pages were read: the Atlas free-tier limits page for the 0.5 GB, the pause and the throttles, and the server limits page for the 16 mebibyte document cap. The page states 0.5 GB.
- evidence:
  - [1] 2026-10-22: 'Free clusters: 0.5 GB' for Data Storage; 'Atlas automatically pauses Free clusters after 30 days of inactivity where there are zero connections'; 10 GB in and out per rolling seven-day period; 100 operations per second; 500 connections; 'You can't configure storage size for Free clusters'
  - [2] 2026-10-22: 'The maximum BSON document size is 16 mebibytes', which is an engine limit and not an Atlas tier limit
- sources:
  - [1] https://www.mongodb.com/docs/atlas/reference/free-shared-limitations/
  - [2] https://www.mongodb.com/docs/manual/reference/limits/

## MongoDB

- `compute/mongodb` | fails-identity | evidence documented | binding write-gated | category managed database
- needs card no, phone no, ID no, institution yes | verified 2026-10-22 (live) | stale on 2026-11-22
- url https://www.mongodb.com/solutions/startups
- free tier: Free MongoDB Atlas credits and Voyage AI tokens across four tiers: Inspire for bootstrapped startups, Grow for early-funded, Innovate for rising VC-backed, and Scale for rapidly scaling. Base-tier eligibility: less than 7 years old, Series A or earlier, building a single software-based product, not previously accepted, with an active company website and LinkedIn profile. | unit: none | metered on: wall-clock | read on: 2026-10-22 | source: [1] | note: one-shot-credit
- per-object cap: not applicable | source: [1]
- retention: not applicable | source: [1]
- egress: not stated | metered: no | source: [1]
- identity: institution | detail: The base tier asks for a company less than seven years old at Series A or earlier with an active website and a LinkedIn profile, and refuses development shops and agencies. The credit value per tier is not published.
- interface: rest
- limits:
  - write-gated (binding): The company-shaped application fails first, and the amount is undisclosed, so even a successful applicant cannot size the benefit before committing to the platform. [1]
- limits note: The company-shaped application fails first, and the amount is undisclosed, so even a successful applicant cannot size the benefit before committing to the platform.
- binding note: the same as limits note
- interface detail: inferred from the tooling the provider documents; not probed
- tooling: not stated
- composes with: Atlas credits at the small-cluster scale are worth tens of dollars a month, so even the top tier is a small absolute benefit beside the compute credits in this class. It composes with a Databricks application, since the two programs overlap on Neon.
- reliability: no reliability event recorded
- notes: One partner landing page, cloud.google.com/mongodb/startups, states up to USD 5,000 in Atlas credits, which contradicts the program page's silence and may be a stale Google-specific co-offer. Recording null rather than either number; the vendor program page is the authority and it publishes none.
- extras:
  - facets.class: startup-credit
  - facets.free_tier_kind: one-shot-credit
- evidence:
  - [1] 2026-10-22: four tiers named Inspire, Grow, Innovate and Scale, each described by funding stage, with no dollar amount published, quoted
  - [1] 2026-10-22: base tier is 'Less than 7 years old and Series A or earlier', a single software product, an active company website and LinkedIn profile, quoted
- sources:
  - [1] https://www.mongodb.com/solutions/startups

## MongoDB Atlas (M0 / Free cluster)

- `storage/mongodb-atlas` | fails-quota | evidence reported | binding none | category paas
- identity a placeholder: card and phone unknown | verified 2026-10-21 (inherited) | stale on 2026-11-05
- url https://www.mongodb.com/atlas
- free tier: 512 MB data storage per free cluster (0.5 GB, uncompressed BSON + indexes); one free cluster per project; 250 projects per Atlas user and 250 per organization | unit: objects | metered on: stored-bytes | read on: 2026-10-21 | source: [1]
- per-object cap: not stated | source: [1] | note: not stated
- retention: not stated | source: [1] | note: not stated
- egress: Data transfer out is not the metered axis on M0 — there is no published egress allowance or overage for Free clusters; the limits that bite are 500 connections and 100 ops/sec. Effectively unmetered for read volume, which makes M0 unusual in this class. | metered: no
- identity: email
- interface: rest
- limits note: 512 MB total is tiny per unit — the whole value is the 250x multiplier. 100 ops/sec, 500 connections, no sharding, no configurable storage. Cluster creation is heavily rate-limited in practice (one free cluster per project, project creation itself throttled).
- interface detail: rest
- tooling: Standard MongoDB drivers (pymongo, motor), `mongosh`, Atlas Data API (HTTPS), `mongodump`/`mongorestore`; GridFS for files >16 MB via pymongo's gridfs
- composes with: 250 projects per user x 1 free cluster each x 512 MB = ~128 GB per Atlas account, and 250 organizations per user gives another axis. GridFS chunks (255 KB default) mean no practical per-object cap beyond the 512 MB total.
- reliability: Free clusters are shared and throttle at 100 ops/sec; a bulk-write workload will be visibly slow rather than failing. Data is 3-node replicated, no vanishing.
- notes: the same as the free tier
- extras:
  - dead_end.failure_class: quota
  - dead_end.why: 512 MB data storage per free cluster (0.5 GB, uncompressed BSON + indexes); one free cluster per project; 250 projects per Atlas user and 250 per organization
  - dead_end.first_recorded: 2026-10-21
- evidence:
  - [2] 2026-10-21: 'M0 | 512 MB | Shared | Shared | Free forever'
  - [1] 2026-10-21: 'Free clusters: 0.5 GB ... all uncompressed BSON documents ... plus the bytes stored in their associated indexes'; 'only one Free cluster per project'
  - [3] 2026-10-21: 'Projects per Atlas User | 250', 'Projects per Atlas Organization | 250'
- sources:
  - [1] https://www.mongodb.com/docs/atlas/reference/free-shared-limitations/
  - [2] https://www.mongodb.com/pricing
  - [3] https://www.mongodb.com/docs/atlas/reference/atlas-limits/

## mongodb atlas free cluster (m0)

- `storage/mongodb-atlas-free-cluster` | fails-quota | evidence reported | binding none | category rejected-provider
- identity a placeholder: card and phone unknown | verified 2026-10-21 (inherited) | stale on 2026-11-05
- url https://www.mongodb.com/docs/atlas/reference/faq/storage/
- free tier: Hard 512 MB, not configurable: 'you are over your space quota, using 513 MB of 512 MB'.. | unit: none | metered on: none | read on: 2026-10-21 | source: [1]
- per-object cap: not recorded | source: [1] | note: not recorded
- retention: not recorded | source: [1]
- egress: not recorded | metered: no
- identity: email
- interface: unknown
- tooling: not recorded
- composes with: the same as tooling
- reliability: the same as the free tier
- notes: the same as the free tier
- extras:
  - dead_end.failure_class: quota
  - dead_end.why: Hard 512 MB, not configurable: 'you are over your space quota, using 513 MB of 512 MB'..
  - dead_end.first_recorded: 2026-10-21
- evidence:
  - [1] 2026-10-21: Hard 512 MB, not configurable: 'you are over your space quota, using 513 MB of 512 MB'..
- sources:
  - [1] https://www.mongodb.com/docs/atlas/reference/faq/storage/