Anthropic's API margins (~92%) top DeepSeek's (~84%) — subscriptions are deliberate loss leaders
bookwormengr · x · 2026-10-31
Based on SemiAnalysis' financial reconstruction, a widely misunderstood picture gets cleared up: Anthropic's subscriptions (incl. Max) are deliberate loss leaders with enormous Opus subsidies — the company breaks even only below 21.3% utilization — while real profit comes from Claude API and Enterprise, with API margins around 92%, above DeepSeek's 84%. OpenAI is the mirror image: its API is the loss leader, hence its dominance on the price-performance Pareto frontier. Charts include SemiAnalysis' Chinese labs' coding-plan API value comparison (GLM Coding Max $108/mo → $1,942; MiniMax Token Ultra $132/mo → $3,088), an Intelligence Index vs cost-per-task scatter, and Claude Team/Enterprise pricing.
More from Venture
- Top Chinese AI creators share 10 monetization lessons: private funnels boost revenue 5-10x — sujingshen · 2026-10-31
- PropFirmMatch busted for cookie stuffing; commissions seized as prize pool — lililashka · 2026-10-31
- Factory acquires typedef_ai to give coding agents an enterprise data context layer — maddiehfaulkner · 2026-10-31
- General Compute raises $100M for inference capacity as CoreWeave stock hits 12-month low — YvesMulkers · 2026-10-31
- Analyst warns of unstable AI compute feedback loop as Kioxia's NAND price warning triggers flash crash — bookwormengr · 2026-10-31
- Blackstone warns Anthropic likely to get high junk credit rating after IPO — Polymarket · 2026-10-31