Anthropic hit a $47 billion annualized revenue run rate in mid-May 2026. This follows a climb from $9 billion at the end of 2025. Claude Code fuels this growth, contributing to a monthly revenue addition that reached $11 billion in March. The coding tool reached a $2.5 billion annualized revenue run rate by early 2026. Anthropic also closed a $65 billion Series H funding round in May 2026, which established a post-money valuation of $965 billion. Between 75% and 85% of total ARR comes from usage-based API business. Consumer subscriptions account for only 5% of total ARR. The company serves over 300,000 business customers. Large accounts paying more than $1 million annually doubled to over 1,000 in the period from February to May 2026. I find the margin improvement the most impressive part of the story. The company’s overall gross margin reached the mid-60% range after it climbed from a negative 94% in 2024. Anthropic’s Chief Financial Officer Krishna Rao stated in a May podcast that net revenue retention reached 500%; among the customers contributing $30 billion in ARR in the first quarter, they had only contributed $2 billion a year ago. Claude Opus 4.7 costs $10 per million input tokens and $37.50 per million output tokens for premium users. Claude Sonnet 4.5 costs $3 per million input tokens and $15 per million output tokens. The coding tool has 29 million daily installs on VS Code and accounts for 7% of all code submissions on GitHub. High computing power demand creates a major bottleneck, as the combined demand for Anthropic and OpenAI could exceed 100 gigawatts by 2030 while current available power sits just over 6 gigawatts.
Databricks grows via enterprise data integration
Databricks surpassed a $7 billion revenue run rate in Q2 2026 and grew over 80% year over year. This follows a $5.4 billion run rate in Q4 2025. The company closed a $5 billion strategic funding round at a $190 billion valuation, led by Coatue. Other investors included Blackstone, MGX, T. Rowe Price, and Sixth Street Growth. AI products, including Lakebase and Genie, exceed a $1.4 billion revenue run rate. Lakebase, a serverless Postgres database for AI agents, exceeded $100 million in revenue run rate. Databricks maintains a net revenue retention rate of approximately 140%. More than 800 customers spend over $1 million annually, and over 70 customers spend over $10 million. The platform supports more than 20,000 organizations, including adidas, AT&T, Bayer, Block, Mastercard, Rivian, and Unilever. Databricks delivers positive free cash flow. Lakehouse revenue surpassed $1.5 billion with over 100% year-over-year growth. Unity AI Gateway provides multi-AI governance and cost controls for enterprises. Databricks is completing investments in the company in excess of $7 billion, which includes $2 billion of additional debt capacity. One significant drawback involves the consumption-based pricing model, which makes cloud spend difficult to predict for users. Mosaic AI workloads introduce variable and spiky expenses that create budgeting difficulties for many engineering teams.
Comparing capital efficiency and scale
Anthropic wins on raw scale while Databricks leads in enterprise integration. Anthropic generates $0.23 in ARR per dollar raised, whereas Databricks generates $0.16 on positive free cash flow. Databricks presents a cleaner institutional entry point because investors pay less per unit of growth. Its adjusted EV/ARR ratio sits at 0.38, compared to Snowflake’s 0.69. Anthropic’s scale relies on the API model where token consumption scales with agentic workflows. Databricks captures value through the data layer with products like Lakebase and Genie. Both companies maintain 140% net revenue retention. Claude Code has a 54% market share in AI-assisted coding. I think the competition between these two models defines the current AI era. Anthropic’s revenue grows as customers adopt more Agentic Workflows, which increases token consumption and corresponding revenue without needing new customers. In contrast, Databricks grows through the expansion of data pipelines and the use of natural-language queries from non-technical users. Databricks’s growth remains tied to the unification of data and AI. Anthropic’s growth remains tied to the sophistication of frontier models. Anthropic’s EBTIT profit margin reached 36% in the second quarter of 2026. You should look at the margin trajectory if you want to see the long-term viability of the API model. Will the infrastructure costs for frontier models eventually erode the high margins Anthropic currently enjoys?
