The serverless
database built
for agents.
One isolated database store, native Model Context Protocol (MCP) endpoints, and hourly metered operations unified under one organization model.
Serverless database for agents
await hub.db("prod").table("memories").put({
hashKey: "agent_7",
data: { status: "active", score: 0.98 }
})Start free. Scale with your agents.
Simple, developer-centric base subscriptions with generous quotas. Expand with granular usage-based metering as your agent traffic grows.
MonkeyDB Free
Unlock essential tools to build and test your first agents.
- 10,000 metered operations / mo
- Org-scoped database namespaces
- Console database browser
- TypeScript SDK integration
- 10,000 DB operations per month included.
- Standard read/write throughput (10 req/s burst limit).
- Strict tenant isolation with automatic org-scoped partitions.
- Standard base62 developer API keys (mk_live_...).
- TypeScript SDK client (@monkeyhub/sdk) for seamless operations.
- Interactive Console DB browser for tables and records.
MonkeyDB Pro
Scale your agents with hourly billing, exports, and MCP endpoints.
- 1,000,000 metered operations / mo
- Nightly S3 exports (JSONL format)
- Full Model Context Protocol (MCP) server
- Priority throughput (100 req/s+)
- 1,000,000 DB operations per month included.
- Priority read/write throughput (100 req/s rate limits).
- Tenant-safe sparse GSIs preventing cross-org query leakage.
- Provisional identity org setup and member approval flows.
- Automated nightly per-org S3 exports in JSONL format.
- Automatic hourly op count metering and usage tracking.
- Native Model Context Protocol (MCP) server integration.
- Exposed markdown SKILL.md endpoint for direct LLM system prompts.
MonkeyDB Ultra
Enterprise-grade database throughput, encryption, and custom SLAs.
- 25,000,000 metered operations / mo
- Custom AWS KMS key encryption
- Dedicated throughput & rate limits
- Python SDK & CLI preview access
- 25,000,000 DB operations per month included.
- High-throughput capacity tailored to production agent volumes.
- Custom AWS KMS key encryption for nightly database exports.
- Dedicated org-wide credential and token replay protection.
- Multi-region export target options.
- Full database audit logging and system health dashboard.
- Early developer access to Python SDK and CLI client.
- Dedicated SLA and tech support communication channels.
What launched in MonkeyDB v1
Model Context Protocol
Expose your databases directly to AI agents via standard MCP endpoints.
Connect Claude, Gemini, or custom LLM frameworks directly to MonkeyDB. A built-in skill endpoint (SKILL.md) and llms.txt index deliver developer docs directly to agents, making tool calling instantaneous and reliable.
Nightly S3 Exports
Never lose a record. Nightly exports save your database to your own S3 bucket.
An EventBridge-triggered serverless flow writes one JSONL object per active table to a KMS-encrypted S3 bucket nightly. Ensures seamless compliance, data ownership, and simple restore procedures.
Tenant-Safe Isolation
Org-partitioned data storage guarantees zero cross-tenant query leaks.
Every single DynamoDB partition key begins with ORG#<orgId> or {orgId}:. Sparse-index hashes automatically embed organization, database, and table boundaries, enforcing tenant safety directly at the storage engine level.
Hourly DB Operation Metering
Hourly usage rollups power automatic quota checks and Stripe billing.
A robust API enforcer monitors database operations and updates a global usage table. Paid tiers scale automatically, while free-tier orgs over quota are gracefully rate-limited with standard 402 HTTP responses.
High-Performance Single Table
MonkeyDB indexes organizational datasets into a shared, performant DynamoDB storage core, using sparse indexes and namespaced partitions to achieve millisecond queries.
Single-Table Structure
MonkeyDB uses a single DynamoDB table model to represent all user databases and tables with maximum storage and throughput efficiency.
Single-Table Mapping
Every record maps to a primary key (PK) and range key (SK). Dynamic indexes structure the records without physical partitions.
PK: ORG#<orgId>
SK: DBTABLE#<db>#<table>
Indices:
i1: { hashField: "status", rangeField: "createdAt" }
i2: { hashField: "agentId" }Sparse GSI Indexing
Declare custom indexes on table creation. Fields are dynamically mapped to sparse Global Secondary Indexes (GSIs) on upsert.
Indexed Queries
GSIs are namespace-encoded (e.g. s:active, n:1700000000) to support string and numeric range operators inside single attributes.
await hub.db("prod").table("memories").query({
index: "i1",
hash: "active",
range: { gte: 1700000000000 },
limit: 50
})Guaranteed Tenant Isolation
Tenant safety is built in. Queries are bounded programmatically by org context, keeping separate customer workspaces completely invisible from one another.
Tenant-Safe Boundaries
Every query pathway is partitioned by organization ID automatically. Cross-tenant leakage is physically impossible.
PK Prefix Enforcement
MonkeyDB automatically prefixes all keys with the request context's org ID (e.g. {orgId}:{db}:{table}:{hashKey}) directly inside DB drivers.
// Resolved request context:
{
orgId: "org_abc123",
actor: { type: "human", id: "usr_7" },
plan: "pro"
}Actor Credentials
Authenticate API requests using standard bearer tokens bound directly to organization and member records.
Context Resolution
The API gateway resolves every Bearer token into a tenant context, matching org details and billing plans before executing queries.
Authorization: Bearer mk_live_xyz... // Returns: X-Request-ID: req_abc Content-Type: application/json
Secure Credentials & Access
Manage developer keys and pending accounts with fine-grained access scopes. Protect organization and billing dashboards from agent key operations.
Secure API Access
Provision and revoke keys programmatically. Credentials never leak org administration or billing scopes.
Key Structure
API keys start with a signature prefix (mk_live_) followed by 32 base62 characters. Replay prevention blocks double-spends.
// Creating key
POST /v1/org/keys
{
"name": "Production Agent Key",
"scopes": ["db:write", "db:read"]
}Member Approval Flows
Signups require explicit validation. Prevent unauthorized human or agent credentials from accessing your tenant.
Approval Pipeline
New signups for existing organization domains are stored in a pending state until an owner account approves the request.
// Pending invitation record
PK: ORG#<orgId>
SK: PENDING#<pendingId>
{
"email": "agent@domain.com",
"role": "agent"
}Backups & Hourly Rollups
Automated backups and transparent resource monitoring. Get full data portability and real-time usage statistics on demand.
JSONL Table Backups
Automated cron triggers export all tables per organization to a secure S3 bucket in standard JSONL format.
Automated Exports
An EventBridge-triggered Lambda executes nightly, writing backup files to a secure, KMS-encrypted S3 bucket. Allows easy data portability.
Bucket: monkeyhub-prod-exports
Key: exports/org_xyz/2026-05-19/db_table.jsonl
Format:
{"PK":"...","SK":"...","d":{"status":"active"}}Hourly Op Rollups
Usage counts are calculated hourly and written to a global DynamoDB table for Stripe billing sync.
Billing Middleware
API enforcers read the hourly usage. If quotas are exceeded, 402 error codes are automatically returned (except in local dev environment).
// Usage Table item:
PK: ORG#<orgId>
SK: USAGE#2026-05-19T12:00:00Z
{
"dbOpsCount": 1420,
"billingSynced": false
}