Baseten raises $1.5B at a $13B valuation as inference infrastructure draws model-lab-scale bets
Baseten closed a $1.5B Series F at a $13B valuation — its fourth raise in 18 months — on the strength of 20x year-over-year revenue growth from self-hosted AI inference infrastructure that lets enterprises run their own frontier models instead of paying OpenAI or Anthropic per call. Inference infrastructure — the recurring cost that compounds with every deployed model query — is now drawing the same equity bets that until recently went exclusively to model labs.
Source: thenextweb.com ↗
Training a model is a one-off, capital-heavy event; inference is a continuous cost that scales with every query a deployed model handles.
Baseten
Why this matters
- → Inference costs now dominate AI production budgets, creating durable recurring revenue.
- → Infrastructure attracts mega-rounds previously reserved for model labs alone.
- → Validates self-hosted models as cheaper alternative to API-dependent deployments.
Inference becomes the prize