pgEdge Starfleet

Not yet independently verified. The launch rests on pgEdge’s own press release and blog post of 28 September 2026; the outside reports found the same day (SD Times, and reposts on PR Newswire and the PostgreSQL news feed) repeat the release rather than test the product. We will update this when it can be confirmed, and remove this note.

pgEdge Starfleet is a PostgreSQL database platform aimed at teams taking AI-built applications from prototype to production, announced by pgEdge, Inc. of Alexandria, Virginia, on 28 September 2026 and available the same day with a 14-day trial that needs no card. pgEdge, an open-source Postgres company, names Bertelsmann, Qube Research & Technologies, Thales and US government agencies among its customers. Each database comes with an MCP server so that coding agents can inspect and change it, a RAG server for retrieval over pgvector, a PostgREST server for access from the browser, and copy-on-write branches for parallel experiments and separate environments; a feature pgEdge calls SafeSession keeps agent connections read-only. The same Postgres can run in pgEdge’s managed cloud, in the customer’s own AWS, Google Cloud or Azure account, on-premises or in an air-gapped site, which is the point for buyers with data-residency rules. Prices on its own page: the open-source toolkit is free to run yourself; the managed service costs $25 a month for a Small database (1 vCPU, 2 GB memory, 25 GB storage), $99 for Large (2 vCPU, 8 GB, 50 GB) and $249 for XL (4 vCPU, 16 GB, 150 GB), with dedicated sizes up to 64 vCPU and 256 GB on request; branches are billed by the hour while they exist, about $0.035 an hour for a Small one. Managed plans include 45-day backups and IP allow-listing, but currently run only in the AWS US East (Ohio) region. The bring-your-own-cloud and self-hosted editions are priced on request and add SOC 2 and HIPAA support, multi-region clusters, FIPS-enabled packages and round-the-clock support.