Supabase at $170M/year growing 221% YoY
Jan-Erik Asplund
TL;DR: From Lovable & Bolt.new to Claude Code & Codex, AI coding agents now spin up 60%+ of new Supabase databases, cementing Supabase as the default backend of AI-generated software. Sacra estimates Supabase hit $170M ARR in May 2026, up from $101M in 2025 and up ~221% YoY. For more, check out our full report and dataset on Supabase.


We previously covered Supabase in October at $70M ARR as vibe coding platforms like Bolt.new & Lovable launched their own platform-native databases, threatening to disintermediate Supabase as the category's default backend.
In February 2026, we covered vibe coding’s backend again, looking at Supabase alongside competitors Convex, Neon, and PlanetScale.
Key points via Sacra AI:
- After becoming the backend of vibe coding (2024-25) as the default API recommended by text-to-app builders like Lovable & Bolt.new for apps requiring database, auth & storage, Supabase's fastest-growing driver of usage has become coding agents like Claude Code & OpenAI’s Codex with database launches up 600% YoY, moving Supabase's distribution from a high-churn prosumer channel to the coding agents every developer uses daily.
- While vibe coding platforms launched their own native backends (Replit Database, Bolt Cloud & Lovable Cloud launched in summer-fall 2025) to capture recurring revenue from app usage vs. high-churn app creation, both Bolt and Lovable white-labeled Supabase’s offering via Supabase for Platforms (while Replit chose Neon), converting nominal competitors into tentpole customers with Lovable at $500M annualized revenue in May 2026, up from $250M in 2025 (+3,471% YoY), valued at $6.6B as of its December Series B for a ~26x multiple and Replit at $525M annualized revenue in April 2026, up 1,775% YoY, valued at $9B for a 17x multiple.
- With registered developers growing from 4M to nearly 10M and paying customers from 100K to 250K+ in 7 months (~2.5% customers to registered users) and ARPC roughly flat at ~$700/year, Supabase is growing primarily on its explosive top-of-funnel volume rather than account expansion, with Sacra estimating Supabase hit $170M ARR in May 2026, up from ~$101M at the end of 2025 (+250% YoY), valued at $10.5B as of its June $500M Series F for a ~60x multiple.
- Supabase's direct competition comes from opinionated backend startups like Convex ($36.4M raised, A16Z) and pure-play Postgres providers like PlanetScale ($105M raised, A16Z), which focuses purely on winning as the most reliable & performant database, competing with Supabase both by offering a graduation path for scaling customers off of Supabase and by powering databases as infrastructure for direct competitors like Convex.
- With AI & agents dramatically accelerating app creation, data querying and data exhaust, database companies like Supabase (composable app DB), turbopuffer (vector DB) at $100M ARR in March up from $75M in 2025 (+2,400% YoY) and ClickHouse (analytics DB) at $250M ARR up from $160M in 2025 (+256% YoY) have been the target of acquisition (see Neon’s $1B acquisition by Databricks) by downstream DB companies going after application DBs and have been acquisitive themselves to compete as a bundle with Supabase looking to move downstream with analytics (Analytics Buckets), vector storage (Vector Buckets), and ETL.
For more, check out this other research from our platform:
- Supabase (dataset)
- Vibe coding's backend
- Supabase at $70M ARR growing 250% YoY
- Vibe coding index
- Replit (dataset)
- Replit passes $500M/year
- Lovable (dataset)
- Bolt.new (dataset)
- Lovable vs Bolt.new vs Cursor
- ClickHouse (dataset)
- Product manager at Firebolt on on scaling challenges and ACID compliance in OLAP databases
- AI program manager at AstraZeneca on running self-hosted ClickHouse
- Turbopuffer (dataset)
- $100M/year PostHog of vector databases
- Databricks (dataset)
- Databricks at $4.8B ARR
- Databricks at $4B ARR growing 50% YoY