Databricks says its annualized revenue rose 80%+ YoY to $6.9B, up from $5.4B in Q4; CEO Ali Ghodsi says AI agent usage is increasing costs and lowering margins
Databricks has a unique role in the AI boom. Revenue continues to soar as businesses swarm to the company's data analytics tools.
Context & Ripple Effects
Databricks’ reported annualized revenue has climbed from $2.4B in mid-2024 to $3.7B in mid-2025, $5.4B in the January 2026 quarter, and now $6.9B. Its customer base had already shown signs of deep enterprise adoption, including nearly 50 customers spending more than $10M annually as of early 2025.
That growth has been accompanied by sharply rising private-market valuations and new financing, reaching a $134B valuation in the company’s February financing. The new disclosure adds a constraint to the growth narrative: heavier AI-agent use is also making the platform more expensive to operate.
First-order effects
- Databricks is converting AI-related demand into a faster revenue run rate, reinforcing its position with enterprises using data analytics tools for AI workloads.
- Higher AI-agent usage directly raises Databricks’ operating costs and compresses margins, creating a near-term trade-off between serving expanding usage and preserving profitability.
Second-order effects
- Databricks will face pressure to improve the efficiency of agent workloads or adjust commercial terms so that high-cost usage does not outpace the revenue it generates.
- Enterprise customers may receive greater scrutiny of agent usage and costs as providers seek to distinguish valuable recurring workloads from computationally expensive activity with weaker unit economics.
Third-order effects
- If agent adoption continues to grow faster than serving efficiency improves, AI infrastructure and data-platform competition will increasingly turn on cost control and workload monetization, not just top-line growth.
- The pattern suggests that enterprise AI’s value chain may concentrate around platforms able to pair proprietary data workflows with sustainable inference economics; whether margins recover depends on usage efficiency and pricing discipline.
The trend: Enterprise AI is moving from early adoption toward an economics test in which rapidly growing agent usage must be matched by durable unit margins.