The $30B figure is extraordinary. Not just because it’s large, though it is, but because of the velocity it implies. Bloomberg, Reuters, and the Financial Times reported the March 2026 figure alongside the December 2025 baseline of approximately $9B. That’s a more than 3x increase in approximately three months. Revenue growth at that rate doesn’t happen without a structural driver.
The attributed drivers are Anthropic’s agentic tool suite, Claude Code specifically, and the Mythos model, though this attribution reflects reporting characterizations rather than a verified revenue breakdown by product. Claude Code has been publicly confirmed as an Anthropic product. Mythos is a real model: Epoch AI has independently evaluated Claude Mythos and confirmed it exceeds current compute thresholds for Sovereign-Class AI designation, a finding with regulatory implications TJS has covered in its technology pipeline. What portion of the $30B ARR is attributable to either product specifically remains unconfirmed from primary sources.
The valuation rejection is the strategic signal that contextualizes the revenue number. Anthropic has reportedly declined offers that would have valued the company at more than $800B. Declining high-valuation investment at rapid growth isn’t unusual for companies with strong revenue visibility, but it’s a meaningful statement about capital strategy. Taking investment at $800B+ requires a credible path to returns at that valuation. Declining it says Anthropic believes it can get there on existing revenue trajectory without the dilution or governance strings that come with a round at that scale.
This is a different posture than OpenAI’s. OpenAI reportedly closed a $122B round in March 2026 at approximately $852B post-money, a round that came with the investor scrutiny TJS is covering in this cycle’s companion brief. Anthropic is choosing not to be in that position. Whether that’s discipline or missed upside depends on how the next twelve months develop.
The competitive comparison matters for enterprise buyers making vendor decisions right now. According to reporting, Anthropic’s March 2026 revenue pace appears to exceed OpenAI’s reported February 2026 figures, though both sets of numbers rely on reported sources rather than primary filings, and the comparison should be treated as indicative rather than confirmed. What the trajectory does confirm is that Anthropic is no longer a smaller alternative to OpenAI in enterprise AI procurement terms. It’s a competitor with comparable or greater revenue velocity.
What to watch: The fall 2027 compute threshold designation for Mythos has regulatory implications that extend beyond the revenue story. If Mythos carries a Sovereign-Class designation under applicable AI governance frameworks, enterprise buyers using it may face documentation and compliance obligations that don’t currently apply to other Anthropic models. Watch the regulatory pillar for that thread. On the financial side, watch whether Anthropic’s Q2 2026 revenue sustains this trajectory or reflects seasonal enterprise contract timing. A single quarter’s run-rate, even at $30B, is a data point, not a trend until it holds.
The TJS read: Two things are simultaneously true. Anthropic’s revenue velocity is real and its valuation rejection represents a coherent capital strategy. And the $30B figure carries enough uncertainty in its sourcing that it should inform analysis, not replace it. Treat it as a strong signal, not a confirmed fact.