Hopsworks 5.0

Hopsworks 5.0 is a sovereign AI Lakehouse that combines a feature store, model registry, and vector database with embedded coding agents (Claude Code and Codex) and a built-in terminal, so teams can go from raw data to production pipelines in one platform. It targets data and ML teams in regulated, sovereignty-sensitive organizations building batch, real-time, and agentic AI systems.

Announced on June 30, 2026, Hopsworks 5.0 is described by the company as its largest release to date, introducing the 'Coding Data and AI Stack' that embeds a full coding agent and terminal (pre-loaded with Claude Code and Codex) directly inside the platform so developers can describe what they want to build and have the system handle ingestion through to deploying production dashboards and inference pipelines. At its core is an AI Lakehouse that unifies a feature store, model registry, and vector database across the full lifecycle from data ingestion to batch, real-time, and agentic inference. The 5.0 release adds native SQL querying via Trino, a dashboarding layer via Apache Superset, 'The Wizard' guided pipeline setup, Platform Intelligence that uses LLMs to configure data sources, column-level access control, model imports from Hugging Face, real-time transformations via RonDB, and expanded connectors to Databricks, Snowflake, BigQuery, MongoDB, and SAP HANA. As a European-founded sovereign AI platform, Hopsworks runs as SaaS, on-premises, or in air-gapped deployments (with GPU and distributed training via Ray and Spark), serving organizations such as Zalando that require data sovereignty and regulatory compliance.

Category

Data & Analytics

Pricing

Contact for pricing

Target Market

Data Scientists, ML Engineers, Data Engineers, CTOs, Heads of AI

Founded

2018

Headquarters

Stockholm, Sweden

• ✓
  Embedded coding agents

  A full coding agent and terminal built into the platform, pre-loaded with Claude Code and Codex, to build pipelines from a natural-language description.

• ✓
  AI Lakehouse core

  A unified feature store, model registry, and vector database covering batch, real-time, and agentic inference.

• ✓
  Native SQL and dashboards

  Run exploratory SQL via Trino and build shareable dashboards via Apache Superset without leaving the platform.

• ✓
  Sovereign and air-gapped deployment

  Runs as SaaS, on-premises, or air-gapped for organizations with data-sovereignty and regulatory requirements.

• ✓
  Broad data integration

  Connectors to Databricks, Snowflake, BigQuery, MongoDB, and SAP HANA, plus model imports from Hugging Face and real-time transformations via RonDB.

• •
  Real-time ML feature serving

  Manage online and offline features with sub-second retrieval for production models.

• •
  Agentic AI development

  Use embedded coding agents to build data and inference pipelines end to end inside the platform.

• •
  Sovereign AI for regulated sectors

  Deploy on-premises or air-gapped to keep sensitive data within compliance boundaries.

Pricing is not published in the launch materials; Hopsworks offers SaaS, on-premises, and air-gapped deployment options.

This page was written from 2 sources, 1 on domains other than hopsworks.ai.

1. 1.globenewswire.com — releasing hopsworks 5 0 introducing the coding data and ai s
2. 2.hopsworks.ai — newsvendor