Anthropic discloses Q3 revenue of $16.8B to investors, sharply cuts distillation usage

At a Wednesday investor event, Anthropic disclosed revenue of about $4.8B in Q1, $11.5B in Q2, and $16.8B in Q3, nearly 50% quarter-over-quarter growth. In Q3, every $1 of compute spend yielded $1.70 in revenue, with an adjusted operating margin of 11% excluding stock-based compensation. Anthropic also tightened "distillation"—competitors training their own models on a more advanced model's outputs. The company told investors distillation contributed 5% to 7% of Q2 revenue and was "substantially" curtailed in Q3. By that math, unnamed labs spent over $500M in Q2 distilling Anthropic models. For teams relying on Claude outputs for training or synthetic data, the pullback limits access to high-quality generation and forces a shift to proprietary data or a re-evaluation of training costs; developers building on the Claude API may also need to revisit call strategies and budgets as the platform reduces supply to peers.