---
title: "OpenAI's agent bill now runs at twice its researchers' cash pay"
description: "OpenAI's own report prices agent use at twice its researchers' cash pay while internal deployment is 2 percent of research compute, and the number lands the week Anthropic lists and OpenAI settles for half its raise. The bill stacks inference on the chip budget, and holders pay it."
publisher: "The Inference"
section: "Model"
published: 2026-11-19T16:39:15.505Z
modified: 2026-11-19T16:39:15.505Z
canonical: https://theinference.org/article/openai-s-agent-bill-now-runs-at-twice-its-researchers-cash-pay
language: en
keywords: "OpenAI, Anthropic, Nvidia, AI Agents, AI Research"
---

# OpenAI's agent bill now runs at twice its researchers' cash pay

> OpenAI's own report prices agent use at twice its researchers' cash pay while internal deployment is 2 percent of research compute, and the number lands the week Anthropic lists and OpenAI settles for half its raise. The bill stacks inference on the chip budget, and holders pay it.

[OpenAI](https://theinference.org/markets/companies/openai) has put its own research loop on the record, and the invoice is the news. A [report published on November 18](https://openai.com/index/research-acceleration-jagged-frontier), "Research acceleration: The jagged frontier," counts what the lab's coding agents cost and do [1].

The median researcher burned $2,400 of tokens a day in October at API list prices, up 62% in a month; the 90th-percentile researcher burned $23,000, up 105% from $11,000 [1]. A heavy user's habit annualizes to roughly $8 million at public rates [1]. The median answer across 101 researchers surveyed: they would trade 40% of team headcount, or 30% of team compute, to keep the agents for three months [1].

## One line prices the whole trade

The report's quietest line is the one that prices the whole trade: as of September 2026, the API-equivalent value of agent use across OpenAI's research organization was about twice the researchers' non-equity compensation, and equity is 70 to 80% of target pay [1]. If equity is three-quarters of the package, cash pay is roughly a quarter of it, so the agent bill works out near half of all-in research wages. Those dollar figures are list prices, what an outside customer would pay, so they mark a ceiling on OpenAI's actual cost; the true number is not disclosed [1][2].

## A labor line, not a chip line

Yet this spend does not compete with the chips investors finance. Internal use of Codex, ChatGPT and the API accounts for just 2% of total research compute [1]. Agent research is inference stacked on training. That is the demand picture [Nvidia](https://theinference.org/markets/companies/nvidia) described on [its November 18 call](https://www.healthcareinfosecurity.com/nvidia-projects-ai-growth-beyond-2027-despite-ceo-exit-a-33436): OpenRouter token volume up 35-fold year over year, agentic workloads about 80% of tokens, [token prices down about 60% since May](https://theinference.org/article/openai-s-new-model-runs-tasks-about-nine-times-cheaper-than-claude-opus-5-5-the-model-in-anthropic-s-s-1) while GPU rental rates rose [4].

*Chart: **The heaviest users more than doubled their daily agent bill in a month** Column chart: daily agent token spend per OpenAI researcher at API list prices, September versus October 2026. Median $1,500 to $2,400; 90th percentile $11,000 to $23,000.*

*Daily agent token spend per researcher at API list prices, September versus October 2026, in US dollars. Source: OpenAI, Research acceleration: The jagged frontier, November 18, 2026. [1]*

And the uplift is jagged. Agents finished 46% of two-to-four-hour tasks unaided in October, up from 23% in March, but only 25% of inference and serving code [1][6].

*Chart: **Agents finish monitoring jobs, not the code that ships models** Bars chart: zero-intervention agent success rates by task type, from 79 percent for monitoring training runs down to 25 percent for inference and serving code.*

*Share of tasks agents completed with no human intervention, by task type, in percent. Source: OpenAI, Research acceleration: The jagged frontier, November 18, 2026. [1]*

## Holders pay for it, researchers carry the risk

OpenAI's holders pay the burn; this landed the week the lab [raised $15 billion at a $1.25 trillion valuation](https://247wallst.com/investing/2026/11/13/openai-raised-only-15-billion-half-what-it-wanted/), [half what it first sought](https://theinference.org/article/anthropic-told-investors-it-turned-its-first-profit-as-openai-races-to-raise-30-billion) [3], while [Anthropic](https://theinference.org/markets/companies/anthropic) [priced its IPO at $83](https://cryptobriefing.com/anthropic-nasdaq-bell-record-ant-ipo/) and debuted on Nasdaq today at about $1.5 trillion fully diluted [5]. Sellers of inference, from Nvidia down to the neoclouds, capture it. Researchers carry the other exposure: their own median answer reprices their colleagues' jobs.

OpenAI's counterpoint is fair: the headcount figure is a hypothetical preference, humans remain central, and Safety and Alignment workweeks grew 77% from January to September [1]. Watch the next report's ratio. If agent spend keeps doubling against a wage bill it already matches, the question stops being whether agents accelerate research and becomes what the researchers are for.

## Takeaway

OpenAI's own November 18 report values its researchers' agent use at about twice their non-equity pay, yet internal agent work is only 2% of research compute, so the bill is inference stacked on the chip budget and the holders pay it.

## Sources

1. [OpenAI, Research acceleration: The jagged frontier, 18 November 2026](https://openai.com/index/research-acceleration-jagged-frontier)
2. [Crypto Briefing, OpenAI's top researchers now burn $23,000 in AI tokens a day, 18 November 2026](https://cryptobriefing.com/openai-researchers-23000-daily-token-usage/)
3. [24/7 Wall St, OpenAI Raised Only $15 Billion, Half What It Wanted, 13 November 2026](https://247wallst.com/investing/2026/11/13/openai-raised-only-15-billion-half-what-it-wanted/)
4. [HealthcareInfoSecurity, Nvidia Projects AI Growth Beyond 2027 Despite CEO Exit, 18 November 2026](https://www.healthcareinfosecurity.com/nvidia-projects-ai-growth-beyond-2027-despite-ceo-exit-a-33436)
5. [Crypto Briefing, Anthropic rings Nasdaq opening bell ahead of record ANT IPO, 19 November 2026](https://cryptobriefing.com/anthropic-nasdaq-bell-record-ant-ipo/)
6. [Digital Today, OpenAI researchers' AI agent use surges, study shows, 19 November 2026](https://www.digitaltoday.co.kr/en/view/128216/openai-researchers-ai-agent-use-surges-study-shows)


---

OpenAI's agent bill now runs at twice its researchers' cash pay — The Inference. Canonical: https://theinference.org/article/openai-s-agent-bill-now-runs-at-twice-its-researchers-cash-pay
