Anthropic is to disclose new revenue numbers at pre-IPO investor day today, CNBC reports citing sources

Pre-IPO investor days of this kind follow a familiar sequence: a curated disclosure of revenue scale, growth rate and unit economics designed to anchor private-markets pricing ahead of a listing, with the numbers framed by management rather than audited for the tape. Episodes of this sort have tended to set the valuation debate through the run-rate figure and its composition, the split between enterprise contracts, API consumption and seat-based licensing being the line that determines which public comparables the deal gets priced against. For AI labs specifically, the disclosures that have moved sentiment are revenue concentration among a small set of hyperscaler and platform partners, the share of revenue tied to compute commitments flowing back to investors, and the burn implied by training spend against gross margin. The tell worth watching is whether the figures are actuals or annualized run-rates, a distinction that has historically separated credible pre-IPO narratives from promotional ones, and whether the disclosed growth justifies or stretches the most recent private round pricing. Follow-ons are syndicate chatter on the IPO timetable, any update to the compute and cloud commitments that structure the company's cost base, and read-across to listed AI-infrastructure names whose revenue lines are the mirror image of this spend.