By Supabase just raised $500 million at a $10.5 billion valuation. But the funding number is almost secondary to what actually drove it: more than 60% of new databases on Supabase are now provisioned not by developers, but by AI agents. Database launches grew 600% in a single year. The headline is a funding round. The real story is that AI tools have become the primary infrastructure operators on one of the fastest-growing developer platforms on the planet.
The Numbers That Explain the Valuation
Supabase’s Series F — led by Singapore sovereign wealth fund GIC, with participation from Stripe, Salesforce Ventures, Accel, and Y Combinator — closed at a $10.5 billion post-money valuation. That’s more than double the $5 billion the company was worth just eight months ago, when it closed its Series E. Total funding now exceeds $1 billion. The company has 10 million developers, 250,000 customers, and 350 employees.
CEO Paul Copplestone did not bury the lead. In his post announcing the round, he was direct: “Claude Code is the largest contributor since the start of the year. Agents are now deploying the majority of databases on our platform.” This is not a vague gesture toward AI adoption. More than 60% of new database launches are created by AI tools, with Anthropic’s Claude Code driving more of them than any other single source. The vibe coding wave — developers describing intent to AI rather than hand-writing every line — is not a trend anymore. It’s Supabase’s core growth engine.
Why AI Agents Default to Supabase
This did not happen by accident. Supabase is built on Postgres, which means AI coding agents can write standard SQL against it without needing specialized knowledge. Its REST and GraphQL APIs are auto-generated from the schema — agents can provision a database without manually configuring any API layer. Auth, storage, and realtime are all part of the same surface, so an AI building a full-stack app hits one platform instead of wiring together five services. And because Supabase is open source, it’s well-represented in the training data LLMs were built on.
The result: when a developer opens Claude Code or Cursor and says “build me a task app with user accounts,” the generated stack almost reflexively reaches for Supabase. The platform made itself legible to machines. That turned out to be worth several billion dollars.
Multigres: The Missing Bridge to Production
The funding announcement came alongside a product release: Multigres v0.1 Alpha, an open source horizontal scaling layer for Postgres. Supabase describes it as “Vitess for Postgres” — a reference to the sharding and scaling infrastructure YouTube built for MySQL at massive scale, now ported to the Postgres ecosystem.
Multigres brings sharding, zero-downtime migrations, automatic failover, context-aware connection pooling, and a Kubernetes operator, all under the Apache 2.0 license. And the timing is not subtle: if AI agents are creating the majority of databases on your platform, many of those databases will eventually need to handle real production load. Supabase is closing the gap between “an AI scaffolded this” and “it can survive a traffic spike.” The stated ambition is to help apps scale “up to the size of OpenAI or even larger.”
What This Actually Means for Developers
Supabase’s trajectory is a proxy for a structural shift in how software gets built. In 2026, AI coding tools are daily-active for 72% of developers, and 41% of global code is now AI-generated, according to TechCrunch. The platforms that win in this environment are the ones machines can operate easily — and Supabase, perhaps more by deliberate design than accident, became exactly that.
The question worth sitting with: if an AI agent provisioned your database, do you know what it created? Multigres addresses the scaling side. The operational side — monitoring, backup ownership, schema drift as agents evolve the app — is still largely on the human in the loop. AI agents are provisioning infrastructure faster than most teams have figured out how to own it at 2 AM when something breaks.
None of that diminishes what Supabase built. Doubling to $10.5 billion in eight months, fueled by AI agents treating your platform as the default backend, is a remarkable outcome. For developers already on Supabase, the practical message is clear: the platform just got a significant scaling upgrade and a large cash injection to keep building. For those evaluating backends for AI-heavy projects, the numbers speak plainly — the AI-native default has become Supabase, and the official Series F announcement makes clear the company intends to hold that position.
