# MongoDB Atlas Vector Search pricing (2026): plans, cost, and compliance

**Entry price:** $0 (free)/mo (M0 Free). Among the 7 vector databases we track it is the cheapest; median entry $0/mo.
**Domain:** mongodb.com
**Free tier:** Yes
**Compliance:** HIPAA Yes · SOC 2 Yes · GDPR Yes
**Integrations:** LangChain, LlamaIndex, OpenAI, Haystack, Microsoft Semantic Kernel, AWS, Spring AI, Voyage AI
**Best for:** Teams already on MongoDB Atlas wanting vector search without a separate DB; Regulated enterprises (healthcare, government) needing HIPAA and FedRAMP; Applications combining document, operational, and semantic search in one query
**Price tracked since:** June 2026

## Plans

| Plan | Monthly | Annual | Billing | Notes | Source |
| --- | --- | --- | --- | --- | --- |
| M0 Free | $0 | — | flat | 512 MB storage, shared RAM/vCPU. Vector Search included. 100 ops/sec max. Free forever. | mongodb.com |
| Flex | $8 | — | custom | From ~$8/mo, usage-based up to ~$30/mo. 5 GB storage, shared resources, 0–500 ops/sec. Vector Search included. Hourly billing. | mongodb.com |
| Dedicated M10 | $58 | — | flat | ~$58.40/mo (hourly billing). 10–128 GB storage, 2 GB RAM, 2 vCPUs. Vector Search included on cluster. Dedicated Search Nodes add separate cost: High CPU S20 at $0.12/hr (~$87.60/mo per node, minimum 2 nodes recommended). | mongodb.com |
| Search Node S20 (High CPU) | $88 | — | flat | $0.12/hr × 730 hrs ≈ $87.60/mo per node. 4 GB RAM, 106 GB storage, 2 vCPUs. Dedicated Vector Search node, minimum 2 nodes for HA. Add-on to existing Atlas cluster. | mongodb.com |
| Search Node S40 (High CPU) | $350 | — | flat | $0.48/hr × 730 hrs ≈ $350.40/mo per node. 16 GB RAM, 426 GB storage, 8 vCPUs. Dedicated Search Node for Vector Search. | mongodb.com |
| Enterprise | Quote-only | — | custom | Custom enterprise pricing for large-scale deployments. Includes dedicated support, SLAs, custom data residency. Contact MongoDB sales. | mongodb.com |

**Key features:** Unified operational + vector data in one platform — no sync required; Auto-generates and syncs vector embeddings via Voyage AI integration; Hybrid queries combining vector search with SQL-style filters, geo, aggregation, and graph lookups; Dedicated Search Nodes for workload isolation (HNSW and ENN algorithms); FedRAMP Moderate authorized (Atlas for Government); FIPS 140-2 encryption.

**Plan limits:** M0 Free: 512 MB storage, shared resources, 100 ops/sec max, free forever; Dedicated Search Nodes require minimum 2 nodes for high availability; Vector Search supports up to 4,096 embedding dimensions.

Sources: 1 · Last verified 2026-06-14
Provenance: https://www.mongodb.com/pricing accessed 2026-06-14. Vector Search is a feature of Atlas — it runs on dedicated Search Nodes (S20–S80 tiers) separate from database nodes, or on standard Atlas clusters. Search node hourly rates confirmed: High CPU S20 at $0.12/hr (~$87.60/mo), up to High CPU S80 at $3.26/hr (~$2,379.80/mo). M0 Free cluster ($0) supports Vector Search. Vector Search is not separately priced — it is included with cluster cost. HIPAA not mentioned on pricing page.
Canonical: https://frontdeskreview.com/software/vector-databases/mongodb-atlas-vector/
Cite: Per The Front Desk Review, MongoDB Atlas Vector Search's entry plan was $0 (free)/mo as of 2026-06-14 — https://frontdeskreview.com/software/vector-databases/mongodb-atlas-vector/
License: CC BY 4.0. Price history: https://frontdeskreview.com/data/price-history.jsonl