← Home
The Twenty Minute VC · July 4, 2026 · 1h 8m

Open Models vs Frontier Models: Who Actually Wins? | The $100,000 Token Budget Every Engineer Will Need | Why Forward-Deployed Engineers Are the Future of Enterprise AI

Clay Bavor, Co-Founder of Sierra (a $15.8B enterprise AI company serving 40% of Fortune 50), discusses the future of frontier vs. open models, the emerging $100K token budget constraint for engineers, and why forward-deployed engineers are becoming critical infrastructure for enterprise AI. Sierra has raised $1.5B+ and surpassed $150M ARR, making it one of the fastest-growing enterprise software businesses in history.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Novel

The $100,000 Token Budget: A Hard Constraint on AI Development24:00
Every engineer will soon face a $100K token budget — a ceiling on the total tokens they can consume in development and deployment, forcing hard tradeoffs between model capability, latency, and cost.

Highlights

Building AI for 40% of Fortune 5020:30
Sierra's rapid penetration into Fortune 50 companies reflects both strong product-market fit and the urgent enterprise need for AI-native operational tools.

Editorial

Frontier AI Demand Will Be Unlimited00:00
Bavor argues that demand for frontier AI models will never plateau — companies will always want access to the most capable models available.
Open Models vs Frontier Models: A False Dichotomy08:00
Bavor suggests the market won't simply split into open vs frontier camps — instead, different companies will use different models strategically based on their specific use case and cost constraints.
Forward-Deployed Engineers as Competitive Advantage37:00
Bavor argues that forward-deployed engineers — embedded with customers, understanding their specific operational problems — are becoming critical infrastructure for enterprise AI success.

Misc

Sierra is valued at $15.8B and serves 40% of Fortune 50 — extraordinary market penetration for an AI-native company
The $100K token budget constraint is becoming real for every engineer — a hard ceiling on what's computationally feasible
Forward-deployed engineers (embedded with customers, understanding their specific problems) are becoming a competitive advantage
Was this useful?