DagsHub
Manage multimodal datasets, annotations, experiments, and models in one open-source-friendly AI platform.
About DagsHub
DagsHub brings multimodal dataset work, experiment tracking, and model management into one platform for AI developers and teams. Projects can keep code, data, models, pipelines, experiments, and annotations together, so changes remain connected to the surrounding project context. The platform supports data and code versioning, dataset curation and annotation, notebook versioning and diffs, CI/CD/CT integration, interactive pipelines, and MLflow-compatible experiment tracking. Its workflow is based on familiar open-source tools and formats, including Git, DVC, MLflow, and Label Studio. The Individual tier is intended for small AI applications and includes public repositories, limited private-project collaboration, 20GB of DagsHub storage, and up to 100 tracked experiments in private repositories. Team adds private-repository, multimodal annotation, connected-storage, RBAC, priority-support, and larger-data capabilities. Enterprise extends the platform for high-scale workloads with deployment to customer clusters, VPC or air-gapped installation, enterprise identity controls, OpenShift compatibility, organizational resource controls, and an enterprise SLA.
Best for
Best use cases
Centralize code, datasets, annotations, experiments, and models in one project
Adopt Git, DVC, MLflow, and Label Studio-compatible workflows without replacing familiar formats
Run large-scale AI data workflows with customer-cluster deployment or private installation
Compare before choosing
The free tier limits private-project collaboration to two collaborators and private experiment tracking to 100 experiments
Team expands private collaboration and data capacity but requires paid-plan evaluation with sales
Enterprise adds private deployment and organizational controls, with pricing available only by custom quote
Features
Curate and annotate multimodal datasets
Connect data sources, query and visualize vision, audio, and LLM datasets, then manage annotation work in the same project context.
Track model and prompt experiments
Record and compare experiment runs with MLflow-compatible tracking while keeping results connected to code, data, and collaborators.
Version AI project assets
Manage versions of code, data, notebooks, models, and pipelines together to preserve lineage across the project lifecycle.
Pros
- Combines datasets, annotations, experiments, models, and project code in one workflow, reducing context switching
- Works with Git, DVC, MLflow, and Label Studio-compatible workflows, which lowers migration friction for teams using these tools
- Provides a free Individual tier with 20GB of storage for small applications and public or non-commercial private work
Cons
- The Individual tier limits private projects to two collaborators, which can block larger team adoption
- Private repositories on the Individual tier are limited to 100 tracked experiments, so active projects may need Team
- Enterprise deployment and controls require a custom quote, making upfront cost comparison difficult
Pricing Plans
Individual
- Unlimited public repositories with unlimited collaborators
- Unlimited private repositories for non-commercial use
- Up to 100 tracked experiments in private repositories
- Up to 2 collaborators in private projects
- 20GB of DagsHub Storage
- Data versioning and lineage
- Community support
Team
- Everything in Individual
- Unlimited private repositories
- Multimodal annotation and auto-labeling
- Connect your own storage
- Label Studio compatible
- Team RBAC
- DagsHub priority support
- Up to 1TB of data or up to 2 million files
Enterprise
- Everything in Team
- Petabyte-scale data management
- Deploy models to your cluster
- Full VPC or air-gapped on-premise installation
- SSO, LDAP, and OIDC RBAC
- OpenShift compatible
- Organizational resource control
- Enterprise SLA and support
* Pricing information is for reference only. Please refer to the latest announcement on DagsHub's official website.