Anthropic Is Pushing New Products, and Big Discounts, to Justify Its Huge Valuation
To justify its ~$1.5T valuation, Anthropic is pushing beyond Claude Code into Claude Cowork, cybersecurity and Claude Managed Agents (still in public beta). Revenue hit $4.7B in Q1 2026, grew 144% in Q2, then 47% to $17B in Q3. Most spending is still on Claude Code; Anthropic says Cowork adoption outpaced Claude Code's first nine months but gave no figures. TENEX.AI says large clients moved 10-20% of Anthropic spend to vulnerability scanning, mostly via their own harnesses rather than Claude Security. Facing OpenAI (up to 40% discounts on Codex) and customers' own multi-model tooling (FedEx, Robinhood, Coinbase, Shopify), Anthropic is offering credits (up to $100,000 per Vertice; up to $2,000 per user for trials), discounts above 30% and 'share-of-wallet' clauses, and since early November it has asked some customers for invoices proving compliance (NPI Financial). One software firm's discount went from ~15% on $1M to ~30% on $3M, with >30% offered above $5M a year. Caveat: customer-side, anecdotal evidence; Anthropic doesn't break out revenue by product.
Why it matters
Shows the quarterly revenue trajectory and the discounting behind Anthropic's growth just as it lists: useful for Dealroom revenue data and for judging revenue quality behind the valuation.