Supabase Raises $500 Million at a $10.5 Billion Valuation Led by GIC

Supabase Raises $500 Million at a $10.5 Billion Valuation Led by GIC

Seven months ago, Supabase was a $5 billion company. As of this month, it is worth more than twice that — and the people doing most of the buying are not people at all.

On June 4, the open-source Postgres development platform announced a $500 million Series F round at a $10.5 billion post-money valuation, led by Singapore's sovereign wealth fund GIC. The raise roughly doubles a valuation the company set only last October, and pushes Supabase's total capital raised past the $1 billion mark. It is one of the clearest signals yet that the infrastructure layer beneath the AI coding boom has become some of the most prized real estate in software.

The deal

The headline numbers are blunt: $500 million in fresh capital, a $10.5 billion valuation, and a lead investor — GIC — that manages the foreign reserves of the Singapore government and rarely chases hype. Every existing backer re-upped, including Accel, Y Combinator, Craft Ventures, Felicis, Peak XV and Coatue. Payments giant Stripe doubled down with its second investment in the company, and Salesforce Ventures came in as a new strategic backer.

What is striking is the velocity. Supabase closed a $100 million Series E at a $5 billion valuation in October 2025. The Series F lands just seven months later at more than double the price — the kind of compression normally reserved for the frontier AI model labs, not the database plumbing underneath them.

The reason, according to co-founder and CEO Paul Copplestone, is that the plumbing has quietly become the product.

"Demand for Supabase is exploding," Copplestone said in the announcement. "Our user base has more than doubled since the Series E and we've seen a 600% increase in databases year-over-year. Claude Code is the largest contributor since the start of the year. Agents are now deploying the majority of databases on our platform."

Read that last line again. The single biggest source of new databases on Supabase is not a human signing up — it is an AI agent, Anthropic's Claude Code chief among them, spinning up backends on a developer's behalf.

Why a database company is suddenly an AI story

Supabase began in 2020 as an open-source alternative to Google's Firebase, wrapping the battle-tested Postgres database with authentication, storage, real-time subscriptions and auto-generated APIs. For years it was a beloved developer tool with a steady, if unglamorous, growth curve. The founder famously pitched the idea as early as 2014 and could not find takers.

Then "vibe coding" happened. As AI assistants and autonomous agents made it possible to build a working application from a few natural-language prompts, those generated apps still needed somewhere to store data, authenticate users and run logic. Supabase turned out to be almost perfectly shaped for that moment: a complete backend an agent can provision in seconds, with a generous free tier and an API surface that large language models already understand well.

The numbers reflect the shift. The company now serves more than 250,000 customers and a developer community measured in the millions. Its "Supabase for Platforms" product — the white-label layer that powers most of the top AI app builders — grew its customer base 370% in the past six months alone. The platform now fields agents spinning up thousands of customer databases per day.

That is the thesis GIC and the rest of the syndicate are underwriting: that as software creation gets automated, value accrues not only to the model makers but to the infrastructure those models reach for by default. If every AI coding tool reaches for Supabase to handle the backend, Supabase becomes a toll road on the entire vibe-coding economy.

What this means for the AI infrastructure layer

The Supabase round fits a broader pattern. Capital is flooding into the unglamorous middle of the AI stack — the databases, vector stores, orchestration layers and deployment platforms that turn a model's output into a running product. Investors have largely accepted that frontier models will be expensive, competitive and partially commoditized. The companies that own the rails between those models and real applications may capture more durable margins.

It also reframes what a "database company" can be worth. Supabase's $10.5 billion mark places it firmly in the upper tier of developer-infrastructure startups, in the company of names like Databricks-adjacent tooling and the serverless platforms. Crucially, much of its moat is open source: the core is freely available, which lowers adoption friction for both human developers and the agents acting on their behalf, while the managed cloud service captures revenue at scale.

Alongside the raise, Supabase previewed Multigres, an open-source horizontal scaling layer for Postgres that brings sharding, zero-downtime migrations and high availability to teams that outgrow a single instance. The release is a tell: Supabase is positioning not just for the long tail of agent-spun hobby apps, but for the largest production workloads — the moment a vibe-coded prototype becomes a real business that needs to scale.

What to watch

The bull case is straightforward. If AI agents keep deploying the majority of new databases, Supabase sits at a chokepoint of the application-building stack, and a $10.5 billion valuation could look cheap in hindsight.

The risks are equally clear. A huge share of new growth is being driven by AI-generated apps, many of which are experiments that may never convert to paying, durable workloads — so the quality and retention behind that 600% database growth matters enormously. Concentration is another flag: if Claude Code is the single largest contributor, Supabase's fortunes are partly tethered to the strategic choices of a handful of AI labs that could, in principle, build or favor their own backends. And the broader backend-as-a-service space is crowded with hyperscalers and well-funded rivals eyeing the same agentic future.

For now, the market has rendered its verdict. A platform once pitched and rejected over a decade ago is now a billion-dollar-plus company whose busiest customers write no code at all. Watch the conversion of agent-spun apps into paying production workloads, and watch whether the AI labs keep reaching for Supabase by default. Those two questions will determine whether this raise marks a peak — or a midpoint.