ML/AI Engineer | OpenAI Global Negotiation Guide

Negotiation DNA: Equity-Heavy + PPU (Profit Participation Units) | Frontier AI Lab | No-Cliff Vesting (Dec 2025+) | +25–40% AGENTIC AI PREMIUM

Region	Base Salary	Equity (PPU/4yr)	Bonus	Total Comp
San Francisco	$265K–$335K	$500K–$900K	—	$390K–$560K
Seattle	$255K–$325K	$500K–$900K	—	$380K–$550K
London	£185K–£235K	£365K–£658K	—	£276K–£400K

Negotiation DNA

OpenAI ML/AI Engineers are at the absolute frontier of artificial intelligence — training and optimizing the world's most capable AI models, building the infrastructure for GPT-5 and beyond, developing multimodal capabilities, and creating the systems that push the boundaries of what AI can do. This is the most sought-after ML role in the world: you work with more compute, better data, and more advanced model architectures than any other company. The Agentic AI Premium (25-40%) applies to engineers building autonomous agent capabilities — tool use, code execution, multi-step reasoning, and the agent frameworks that enable GPTs and Assistants to take actions in the real world. [Source: OpenAI ML/AI Eng Comp 2025-2026]

Level Mapping: OpenAI ML/AI Eng = Google L5 Research Eng = Meta FAIR Eng = Anthropic ML Eng = DeepMind Research Eng

No-Cliff Day-1 Vesting — The Liquidity Revolution

OpenAI eliminated the 1-year vesting cliff in December 2025. New hires now begin vesting equity from Day 1. "As an ML/AI Engineer at the frontier, my contributions to model training and capabilities begin immediately. OpenAI's No-Cliff vesting means my PPUs — potentially $500K-$900K over 4 years — start vesting from my first month. At Google or DeepMind, I'd wait 12 months before seeing any equity, forgoing $125K-$225K. For ML roles where every researcher is being recruited by every AI lab simultaneously, No-Cliff vesting is a decisive advantage."

Global Levers

1. Frontier Model Training: "I train and optimize the world's most capable AI models — GPT-5 and beyond. More compute, better data, more advanced architectures than any other lab. This is the pinnacle of ML engineering."
2. No-Cliff Day-1 Vesting — ML Premium: "ML engineers are the most recruited talent in tech. OpenAI's No-Cliff vesting means $125K-$225K in equity I'd forfeit during a cliff at Google, DeepMind, or Meta FAIR."
3. Agentic AI Premium (25-40%): "I build the autonomous agent capabilities — tool use, code execution, multi-step reasoning — that enable GPTs and Assistants to act in the real world. This is the purest form of agentic AI, and it commands a 25-40% premium."
4. PPU Upside — ML-Driven Valuation: "OpenAI's valuation is fundamentally driven by model capabilities — my work. Every capability improvement I ship directly increases the value of my PPU grants. No other equity is this directly tied to ML engineering output."

Negotiate Up Strategy: "I'm targeting $328K base and $880K PPUs over 4 years for this ML/AI Engineer role with the 25-40% Agentic AI Premium. I train and optimize frontier models at the world's most advanced AI lab, and I build the agentic capabilities that define OpenAI's next chapter. No-Cliff vesting means $125K-$225K in equity I'd forfeit at a cliff competitor. I have competing ML offers from [Google Brain at $520K TC / Anthropic at $525K TC with $335K base / DeepMind at $510K TC]. OpenAI's compute advantage and PPU upside are unmatched, but the grant must reflect frontier ML contribution with the Agentic AI Premium." Accept at $298K+ base and $800K+ PPUs.

Evidence & Sources

• [OpenAI ML/AI Eng Comp — Levels.fyi 2025-2026]
• [OpenAI No-Cliff Vesting — December 2025 Policy Change]
• [Frontier Model Training — GPT-5 and Beyond]
• [Agentic AI Premium — 25-40% for Autonomous Agent Capabilities]
• [PPU Structure — ML-Driven Valuation Upside]