As a software engineer who has monitored LLM viability as a coding assistant, the idea that 90% of Anthropic's lifetime revenue up to the end of Q1, was in Q1 itself, seems completely realistic to me.
And if every model is profitable but they're getting income on a model that costs X while training the one that costs 10X, everything else makes sense too, and this is indeed their highly reasonable claim.
Anthropic is selling a lot of tokens to companies at full price right now, where it's almost certainly very profitable given that Opus costs several times the current going prices for open model inference.