Bases

Bases

Full environments for your agents.


Create and manage isolated per-tenant context instances programmatically. Memory, files, search, and extraction per tenant, user, or app.

Full environments for your agents.


Create and manage isolated per-tenant context instances programmatically. Memory, files, search, and extraction per tenant, user, or app.

Read the docs

// Create a Base
{
  "name": "customer-acme-corp"
}

// Response
{
  "id": "b3fc-2c963f66afa6",
  "name": "customer-acme-corp",
  "type": "base",
  "createdAt": "2026-01-15T08:00:00.000Z"
}
// Create a Base
{
  "name": "customer-acme-corp"
}

// Response
{
  "id": "b3fc-2c963f66afa6",
  "name": "customer-acme-corp",
  "type": "base",
  "createdAt": "2026-01-15T08:00:00.000Z"
}

Find out how Exabase could cut token spend for your agent product.

Book a demo

Find out how Exabase could cut token spend for your agent product.

Book a demo

Find out how Exabase could cut token spend for your agent product.

Book a demo

What are Bases?

Bases are isolated per-tenant context instances you create on demand via the API.

Each Base gets its own storage, memory, search index, and extraction pipeline. Data in one Base is completely separate from data in another.

Use them to give each tenant, user, or agent their own workspace. One API call to create. Everything scoped automatically.

Problem

Stitching together infra for your agents

You want to give your agent a place to work. Not just memory. Not just files. A full environment where it can store state, create documents, search its own history, and pick up where it left off.


Building this yourself means stitching together databases, file storage, vector search, auth, and isolation. That's months of infrastructure work.

Solution

Bases give agents a home

An Exabase Base is a complete environment with storage, semantic search, memory, and more.


Each Base is isolated. Your agent's data is separate from other agents, other users, other apps.


Use it as a secure store for a tenant of your app, or a knowledge base instance for your agent.

Production-ready

Private by design

Security-first

Scalable

Production-ready

Private by design

Security-first

Scalable

Production-ready

Private by design

Security-first

Scalable

How do I use it?

  1. Create a Base

One API call. Give it a name. That's it.

import { Exabase } from "@exabase/sdk";

const api = new Exabase({
  apiKey: process.env.EXABASE_API_KEY,
});

const base = await api.bases.create({
  name: "customer-acme-corp",
});

console.log(base.id); // use this ID for all scoped operations

  1. Use the Base

Pass the Base ID on any API call. Resources, Memory, Search, Extract, and Workers all scope to it automatically.

// Store a file in this Base
await api.resources.createFile(
  { file: createReadStream("./contract.pdf"), name: "Contract" },
  { baseId: base.id }
);

// Search within this Base
const results = await api.resources.search(
  { text: "payment terms" },
  { baseId: base.id }
);

// Add a memory to this Base
await api.memories.create(
  { messages: [{ role: "user", content: "Acme prefers net-60 terms" }] },
  { baseId: base.id }
);

Bonus: Audit via UI

Navigate any Base through the Exabase GUI. See exactly what your agent has stored, created, or modified. Debug and inspect without writing code.

What each Base includes

Every Base is a complete, isolated environment. No setup required beyond creating it.

File storage

Upload and manage files, notes, bookmarks, and documents

Folders and tags

Organize resources into folders, apply tags, rename and move

Deep Search

Multi-modal hybrid search across all resources in the Base

Memory

Per-Base memory that persists across agent sessions

Extract

Process files and URLs into structured data, scoped to the Base

Workers

Autonomous agents that maintain the Base on a schedule

Workspace Explorer

Browse and audit everything in the Base through the GUI

Bases API at a glance

Endpoint

Method

Description

/v2/bases

POST

Create new Base

/v2/bases

GET

Get list of Bases

/v2/bases/{baseId}

GET

Get a Base

/v2/bases

PATCH

Update Base

/v2/bases

DELETE

Delete Base

View full API reference →

Multi-tenancy built in

Bases are designed for multi-tenant applications. Create one Base per customer, per user, or per project. Data isolation is automatic.

Per-customer workspaces

Each customer gets their own Base. Their documents, memories, and search results are completely separate from every other customer.

Per-user agents

Give each user's AI assistant its own Base. Preferences, conversation history, and saved content stay private.

Per-project knowledge bases

Spin up a Base for each project. Research, documents, and tasks stay organized and searchable within their project scope.

Billing rolls up

All usage from Bases is billed to the parent developer account. Your tenants don't need Exabase accounts.

Limits

100 Bases on the free plan. 10,000 on paid plans. Contact us for higher limits.

Bases API | Isolated Cloud Environments for AI Agents | ExabaseA version timeline v1 through v5 with a looping animated arc showing a rollback from v5 to v3, where v3 holds as the active version.v1v2v3v4v5

Filesystem versioning

Checkpoint your agent's Base and storage state. Roll back if something goes wrong. Browse your Base as it existed at any point in the past. Diff to see what changed.


Version control for agent state.


How it works. Every write to a resource is captured. Query any supported endpoint with a snapshotAt timestamp to see the state at that moment, including resources that have since been deleted.


Granularity. Hourly within the last 24 hours. Six-hour windows for 1 to 7 days. Configurable retention.


Available on all paid plans at no additional cost.

Use cases

Multi-tenant SaaS

Give each customer their own isolated workspace. Documents, memories, and search results stay separate. No data leakage between tenants.

Per-user AI assistants

Each user gets a personal Base. Their assistant remembers their preferences, stores their files, and searches only their content.

Per-user AI assistants

Each user gets a personal Base. Their assistant remembers their preferences, stores their files, and searches only their content.

Project workspaces

Spin up a Base per project. Research, documents, notes, and tasks stay organized and searchable within scope. Archive or delete when the project ends.

Project workspaces

Spin up a Base per project. Research, documents, notes, and tasks stay organized and searchable within scope. Archive or delete when the project ends.

Agent sandboxes

Test agents in isolated environments before deploying to production. Snapshot before a risky operation, roll back if needed.

Agent sandboxes

Test agents in isolated environments before deploying to production. Snapshot before a risky operation, roll back if needed.

Customer onboarding

Create a Base per customer, pre-populate it with documents, and set up a Worker to keep it current. Ready-made knowledge base from day one.

Classroom or cohort environments

Each student or cohort gets their own workspace. Course materials are shared. Student work is isolated.

Classroom or cohort environments

Each student or cohort gets their own workspace. Course materials are shared. Student work is isolated.

Compliance-scoped environments

Separate regulated data into its own Base. Apply different retention policies, access controls, and audit requirements per environment.

Ready for scale

Fast

Deliver speed to your users. Infrastructure that won't slow your agent down.

Secure

Encrypted in transit (SSL) and at rest (AES-256). CASA certified.

Reliable

99.9% uptime. Built on Exabase's consumer-grade scaled infrastructure.

Ready for scale

Fast

Deliver speed to your users. Infrastructure that won't slow your agent down.

Secure

Encrypted in transit (SSL) and at rest (AES-256). CASA certified.

Reliable

99.9% uptime. Built on Exabase's consumer-grade scaled infrastructure.

Ready for scale

Fast

Deliver speed to your users. Infrastructure that won't slow your agent down.

Secure

Encrypted in transit (SSL) and at rest (AES-256). CASA certified.

Reliable

99.9% uptime. Built on Exabase's consumer-grade scaled infrastructure.

Why Exabase

One API call to create a full environment

No provisioning databases, search indexes, or storage buckets. A Base comes with everything: files, folders, tags, memory, search, extraction, and workers. Ready immediately.

One API call to create a full environment

No provisioning databases, search indexes, or storage buckets. A Base comes with everything: files, folders, tags, memory, search, extraction, and workers. Ready immediately.

True isolation

Data in one Base is invisible to another. Search only returns results from the current Base. Memory is scoped. Workers operate within their Base only. Isolation is enforced at the infrastructure level, not just by filtering query results.

True isolation

Data in one Base is invisible to another. Search only returns results from the current Base. Memory is scoped. Workers operate within their Base only. Isolation is enforced at the infrastructure level, not just by filtering query results.

Everything scopes automatically

Pass the X-Exabase-Base-Id header and every API call (Resources, Memory, Search, Extract, Workers) scopes to that Base. No per-endpoint configuration.

Everything scopes automatically

Pass the X-Exabase-Base-Id header and every API call (Resources, Memory, Search, Extract, Workers) scopes to that Base. No per-endpoint configuration.

Audit through the GUI

Browse any Base through the Exabase workspace explorer. See every file, note, and memory your agent created. Debug agent behavior by inspecting what it stored. No custom admin panel needed.

Audit through the GUI

Browse any Base through the Exabase workspace explorer. See every file, note, and memory your agent created. Debug agent behavior by inspecting what it stored. No custom admin panel needed.

Snapshots for rollback

Enable filesystem snapshotting and query any Base as it existed at a past point in time. Roll back after a bad agent run. Compare states. Recover deleted resources.

Snapshots for rollback

Enable filesystem snapshotting and query any Base as it existed at a past point in time. Roll back after a bad agent run. Compare states. Recover deleted resources.

Usage rolls up to one account

All Bases bill to the parent developer account. Create thousands of tenant workspaces without requiring each tenant to have their own Exabase account.

Usage rolls up to one account

All Bases bill to the parent developer account. Create thousands of tenant workspaces without requiring each tenant to have their own Exabase account.

Scale to 10,000+ Bases

Free plan supports 100 Bases. Paid plans support 10,000. Contact us for higher limits. Creation is instant regardless of count.

Scale to 10,000+ Bases

Free plan supports 100 Bases. Paid plans support 10,000. Contact us for higher limits. Creation is instant regardless of count.

vs building your own multi-tenant storage

You provision a database, set up tenant isolation, add file storage, integrate a vector search engine, build an auth layer, and create an admin UI. Exabase does all of that in one API call. And it comes with memory, extraction, workers, and a GUI for free.

vs building your own multi-tenant storage

You provision a database, set up tenant isolation, add file storage, integrate a vector search engine, build an auth layer, and create an admin UI. Exabase does all of that in one API call. And it comes with memory, extraction, workers, and a GUI for free.

Part of a full platform

Bases aren't just storage containers. Every Base has access to Memory, Resources, Deep Search, Extract, and Workers. They all share the same infrastructure and scope automatically. No integration work between services.

Part of a full platform

Bases aren't just storage containers. Every Base has access to Memory, Resources, Deep Search, Extract, and Workers. They all share the same infrastructure and scope automatically. No integration work between services.

vs separate Pinecone/Weaviate namespaces

Vector database namespaces give you isolated search. But you still need file storage, memory, metadata, and a way to manage it all. A Base is a complete environment, not just a search index.

vs separate Pinecone/Weaviate namespaces

Vector database namespaces give you isolated search. But you still need file storage, memory, metadata, and a way to manage it all. A Base is a complete environment, not just a search index.

Part of a full platform

Memory

Self-managing advanced memory system.

Bases

Isolated per-tenant instances with version rollback.

Resources

Files, notes, links. Your portable context server.

Deep Search

Sub-document multi-modal hybrid search.

Extract

Structured data from PDFs, websites, images, and more.

Workers

Autonomous agents and self-enriching knowledge bases.

Memory

Self-managing advanced memory system.

Bases

Isolated per-tenant instances with version rollback.

Resources

Files, notes, links. Your portable context server.

Deep Search

Sub-document multi-modal hybrid search.

Extract

Structured data from PDFs, websites, images, and more.

Workers

Autonomous agents and self-enriching knowledge bases.

Works with everything

Model-agnostic. Framework-agnostic.

SDK

TypeScript SDK. A convenient toolkit to move faster.

API

Simple, clean API. Up and running in under a minute.

Works with everything

Model-agnostic. Framework-agnostic.

SDK

TypeScript SDK. A convenient toolkit to move faster.

API

Simple, clean API. Up and running in under a minute.

Works with everything

Model-agnostic. Framework-agnostic.

SDK

TypeScript SDK. A convenient toolkit to move faster.

API

Simple, clean API. Up and running in under a minute.

FAQs

What is Exabase?

Exabase is infrastructure for AI agents. It gives your agents memory, versioned file storage, AI deep search, and context automation through a set of APIs. Store what your agent learns, search inside any content type, and keep knowledge bases current automatically. Built for production use. Give your agent precise context and cut your token spend by up to 81%.

Who uses Exabase?

Developers and teams building AI agents, copilots, and RAG applications. If your agent needs to remember things between sessions, store and retrieve files, search across documents and media, or stay up to date without manual maintenance, Exabase handles that infrastructure so you can focus on your product.

What is an Exabase Base?

A Base is an isolated cloud filesystem instance you create on demand. It comes with its own file storage, folders, tags, memory, search index, extraction pipeline, and workers. All data in a Base is separate from all other Bases.

How do I create a Base?

One API call: POST /v2/bases with a name. The Base is available immediately.

How do I scope API calls to a specific Base?

Pass the X-Exabase-Base-Id header on any request. All operations (Resources, Memory, Search, Extract, Workers) automatically scope to that Base.

Are Bases truly isolated?

Yes. Data in one Base is invisible to another. Search only returns results from the current Base. Memory is per-Base. Workers only access their own Base. Isolation is enforced at the infrastructure level.

How many Bases can I create?

100 on the free plan. 10,000 on paid plans. Contact us for higher limits.

Do my tenants need Exabase accounts?

No. All Bases are created and managed through your developer API key. Usage rolls up to your account. Your tenants interact with your app, not with Exabase directly.

What does each Base include?

File storage, folders, tags, memory, Deep Search, Extract, Workers, and a browsable GUI through the Exabase console. Everything is available immediately on creation.

Can I browse a Base through a GUI?

Yes. Navigate any Base through the Exabase workspace explorer. See every file, note, memory, and folder your agent created. Useful for debugging and auditing.

What are filesystem snapshots?

Snapshots let you query a Base as it existed at any point in the past (up to 7 days). Every write is captured automatically. Use snapshots to roll back after a bad agent run, compare states, or recover deleted resources. Available on paid plans.

How does snapshot granularity work?

Hourly within the last 24 hours. Six-hour windows for 1 to 7 days. Records older than 7 days are pruned. You can set a shorter retention period if needed.

Can I delete a Base?

Yes. Deleting a Base permanently removes all its data: files, memories, search index, workers, and history. This requires a confirmation step via the API.

What happens to the data when I delete a Base?

All data is permanently deleted, including files, memories, resources, and snapshot history. This cannot be undone.

How is billing handled for Bases?

All usage from all Bases rolls up to the parent developer account. You pay one bill. Individual Bases don't have separate billing.

What's the difference between a Base and a workspace?

A workspace is the top-level account container. Bases are instances you create within a workspace to isolate data per tenant, user, or project. Think of the workspace as the parent and Bases as children.

Can I use Bases with Workers?

Yes. Workers run inside a specific Base. They can only access data within that Base. Create a Worker per Base to maintain each tenant's knowledge base independently.

Deciding?

Ask your favourite AI about Exabase:

Cut your token spend and give your agent precise context.

Get started in minutes.

Start for free

Read the docs

Cut your token spend and give your agent precise context.

Get started in minutes.

Start for free

Read the docs