Dolt

Dolt is for teams that want a relational database whose changes can be branched, reviewed and merged. It is especially suitable for data-heavy development workflows that already use SQL and Git-like collaboration, while organizations handling confidential data should plan around DoltHub’s public default.

Overview

Dolt stores tabular data in a SQL database while adding fork, clone, branch, merge, push and pull operations familiar from Git. You can connect to it as you would to MySQL, or deploy it as a versioned MySQL replica. That makes it a stronger fit than a conventional database when the history and review of data changes matter as much as the current rows.

The core Dolt product is free and open source. DoltHub provides a public web collaboration layer, Hosted Dolt supplies managed deployments, and DoltLab runs a DoltHub-style service in your own network or on a development machine.

Key features

Version control for tables: Branches, commit inspection, diffs, SQL conflict resolution and commit reverts give teams a way to isolate and audit data changes. Dataset branching and point-in-time rollback are supported, with tables as the snapshot granularity. This is useful for experimentation and review, but it is a workflow centered on table-level history rather than arbitrary file snapshots.

MySQL compatibility: Existing SQL-oriented applications can connect through the MySQL interface, and Dolt can serve as a versioned MySQL replica. That lowers the barrier for teams whose applications already depend on MySQL conventions.

Web collaboration: DoltHub hosts public data at no charge and includes browser editing and pull-request review. Databases are public by default; Pro users and organizations can make databases private and assign collaborator or organization permissions. Free public hosting is therefore a poor match for proprietary datasets.

Hosted SQL workbench: Every Hosted Dolt deployment includes a web workbench with pull requests, diffs, commit logs, ER diagrams and CSV import/export. It gives reviewers a browser-based control surface instead of requiring every task to happen at the command line.

Connectors and frameworks: Official connector libraries cover JDBC, Perl DBI, MySQL Connector/Python and Go sqlx/database/sql. Officially tested integrations include Django, SQLAlchemy, Prisma, Laravel, Hibernate, GORM, Rails and FastAPI, making Dolt more practical in established application stacks.

Managed operations: Hosted Dolt performs full nightly backups and keeps them for 14 days, supports AWS and GCP private networking, and includes enterprise support. Deployment-console tickets receive an answer from an on-call team member within an hour. DoltHub also states it has SOC2 compliance and publishes its controls and processes on a Trust page.

Pricing

Dolt: free. The open-source plan is the right starting point for developers and teams able to run their own database and accept public DoltHub collaboration or self-host with DoltLab. Its program size is approximately 100 MB, so larger workloads may need another plan or deployment approach.

DoltHub Pro: $5 a month, billed monthly. It includes 100 MB of private data storage free; storage above 5 GB costs $1/GB/month. Pro is the inexpensive route for private DoltHub databases and permissioned collaboration, but the storage allowance and overage price matter for growing datasets.

Hosted Dolt trial instance: $50 a month, billed monthly, with a t2.medium EC2 instance and 50 GB of storage. There is no time limit on the trial instance. This is a practical managed starting point for teams that want the hosted workbench, backups and support without operating infrastructure, but it is not a zero-cost evaluation.

There is no free trial for Dolt itself. Hosted Dolt enterprise arrangements use custom pricing when the trial instance no longer fits.

Platforms

Dolt supports API, Linux, macOS, self-hosted, web and Windows environments. The deployment model is both hosted and self-managed, so teams can keep operations in their own network with DoltLab or use Hosted Dolt and DoltHub. The data scope is tables, with both dataset branching and point-in-time rollback available.

Who it's for

Choose Dolt when SQL access must coexist with reviewable, reversible data work: application teams, data maintainers and collaborators who need branches, diffs and pull requests around tabular datasets. It is also a sensible option for organizations that need MySQL compatibility but want versioned replicas or controlled deployment choices.

Look elsewhere if your data cannot be public and you do not want to pay for private DoltHub storage, or if your primary requirement is a different kind of data or file versioning rather than table history.

Pros and cons

Pros

Git-style database workflow: Branch, merge, diff and revert operations make data changes reviewable instead of opaque.

MySQL-compatible access: Existing MySQL-facing applications and tested framework integrations can connect without abandoning familiar tooling.

Flexible operations: Free open source, self-hosted DoltLab and managed Hosted Dolt cover different infrastructure preferences.

Useful hosted controls: Nightly 14-day backups, private networking, a browser workbench and rapid enterprise support address operational needs beyond the database engine.

Cons

Public by default: Free DoltHub databases expose data unless you move to Pro or an organization with private permissions.

Small included private allowance: Pro includes only 100 MB free, with $1/GB/month charges above 5 GB.

Hosted entry cost: The $50-per-month trial instance is managed and persistent, but teams seeking a free hosted evaluation will not find one.

Table-focused history: Snapshot and rollback granularity is at the table level, which may not match workflows requiring finer-grained object versioning.

Alternatives

Data Version Control Tools is the broader category to browse when you want to compare table, file and dataset-oriented systems.

Weights & Biases is the better choice for machine-learning teams wanting a freemium service with hosted collaboration, mobile access and defined free storage and ingestion quotas.

Apache Paimon suits teams seeking an open-source data-lake platform where connectors and storage plugins can be selected for the environment.

ClearML is preferable when an open-source self-hosted ML platform and a free community tier are the priority.

git-annex is for managing large files while keeping their contents out of Git.

Nile fits users who want local-only data version control with no cloud account.

DataLad is another free, open-source option for Linux, macOS and Windows.

Graviti targets hosted dataset management with a free starter allocation and paid cloud tiers.

Neon is a hosted database alternative with a free plan, project and branch quotas, and usage-based limits.

Verdict

Dolt is the strongest choice for teams that need SQL-compatible tables plus Git-like branching, review and rollback, with the option to self-host or use managed infrastructure. Choose it for versioned data collaboration; look elsewhere when public-by-default sharing, table-level history or the paid private and hosted tiers do not suit your governance or scale.