SurrealDB, the London company behind a multi-model database, raised $23m as an extension of its Series A round. That takes the Series A to $38m and the company's total funding to $44m, including its seed financing.
The extension was announced on 17 February 2026, the same day as the general availability release of SurrealDB 3.0. Tech.eu dated its report 17 February and the company's own blog post carries 18 February. No lead investor was named and no valuation was disclosed.
Chalfen Ventures and Begin Capital are new investors, joining existing backers FirstMark and Georgian. Mike Chalfen, founder of Chalfen Ventures, is joining SurrealDB's board.
“This fresh investment demonstrates a growing level of excitement for our category-defining, developer-friendly database.”Tobie Morgan Hitchcock, co-founder and CEO, SurrealDB
What SurrealDB does
SurrealDB was founded in 2021, according to TechTarget.
SurrealDB describes its goal as building the first AI-native, multi-model database. It combines relational, document, graph, vector, search, geospatial, time-series and key-value data in a single platform, and its website also lists an Agent Memory product for AI agents and applications.
Chief executive and co-founder Tobie Morgan Hitchcock argued in the announcement that AI agents need persistent memory, a unified view of structured and unstructured data, synchronised context and real-time performance at scale. SurrealDB is built in Rust, according to Tech.eu.
The company reported adoption figures as of February 2026: 2.3 million downloads, 31,000 GitHub stars and more than 1,000 forks, along with what it called rapid adoption in production environments around the world.
SurrealDB 3.0
Version 3.0 reached general availability with the funding news. Tech.eu quoted the company as calling it its most stable, high-performance and enterprise-ready version to date.
TechTarget listed the main additions as Surrealism, a control layer for customising business logic including access and version controls; improved vector search and indexing for unstructured data such as text and images; architectural changes for stability and performance; and developer-experience changes such as defining custom API endpoints inside the database. It also reported context graphs and agent memory positioned next to the data.
What the money is for
SurrealDB said it will invest in enterprise-grade reliability and performance, new security and governance features, cloud scalability and new regions, production support and deployment tooling, and a larger global team. TechTarget reported it also plans to accelerate product engineering and improve go-to-market efforts.
Other outlets reported the extension in other currencies, including £17m and €19m, which correspond to the same $23m.
Chalfen said every compute era requires a new database paradigm.
Hitchcock told TechTarget his priorities are maturing the platform for enterprise-scale use, expanding capabilities so customers need fewer outside tools to build AI applications, and continuing to improve the developer experience.
Analyst view
TechTarget noted that venture funding for data management vendors fell sharply after tech share prices dropped in mid-2022, and that vendors attracting investment now tend to stress AI development, as SurrealDB does. It listed Couchbase, Redis, PostgreSQL, AWS, Google Cloud and Microsoft among competitors with multimodel capabilities.
Kevin Petrie of BARC US said the funding reflects the pain enterprises feel as they adopt AI and can help SurrealDB deepen its product. Matt Aslett of ISG Software Research said SurrealDB is in the early stages, but considers it ahead of many established providers in vector search and indexing, native agent memory and support for several data types in one database.
FirstMark and Georgian were already backing the company. The sources do not state a headcount.
Company website: surrealdb.com