ARK Weekly Report: AI Agents Will Continue to Drive the Growth of Anthropic and OpenAI, Grok 4.6 May Lower Frontier AI Cost Curve

On August 23, ARK Invest, led by Cathie Wood, released its weekly market report covering three main themes:

AI agents will continue to drive the growth of Anthropic and OpenAI: Anthropic submitted an S-1 draft to the SEC on June 1 in preparation for a potential IPO and has been engaging with investors to gauge market sentiment. As of the end of May, Anthropic's annual recurring revenue (ARR) had reached $47 billion, starting from about $9 billion in early 2026, growing more than fivefold in just five months; data provider TickerTrends estimates its current ARR could exceed $74 billion. OpenAI's ARR is approximately $41 billion, doubling from $20 billion at the beginning of the year. Together, the two companies' ARR has surpassed $115 billion, exceeding the combined 12-month revenue of SAP, Salesforce, and Adobe, and approaching Microsoft’s productivity and business processes division's annual run rate of about $150 billion. Both companies plan to finance large-scale computing power construction through public markets.

Grok 4.6 may further drive down the frontier AI cost curve: The released Grok 4.6 is designed for programming, agent tasks, and knowledge work, with key parameters including a context window of 500,000 tokens, input costs of $2 per million tokens, and output costs of $6 per million tokens, significantly lower than GPT-5.6 Sol ($5/30) and Claude Fable 5 ($10/50). Its Intelligence Index score is 61, on par with GPT-5.6 Sol, only trailing Claude Opus 5 and Fable 5 by about 1 to 2 points, but at a much lower price. In terms of task costs, Grok 4.6 is about $0.84 per task, positioned at the intelligent-cost Pareto frontier; its long-term agent knowledge work Elo score is 1577, roughly equal to Claude Fable 5's 1574. Grok Bot has also been launched, extending competition from model performance to the agent software layer capable of running tasks for extended periods. As model capabilities improve, cost metrics should shift from "per token price" to "per completed task price"; if training and inference costs continue to decline by 85% and 99.9% annually, respectively, enterprises will deploy agents in more workflows, significantly accelerating AI adoption.

Minimal Residual Disease (MRD) detection proves its utility and continues to scale: Key evidence comes from bladder cancer trials IMvigor010 and IMvigor011, where ctDNA positive patients showed a 50% improvement in median overall survival after using atezolizumab. Last quarter, Signatera sales reached 283,000 (an increase of 34,000 quarter-over-quarter), while all other solid tumor MRD detections totaled only 41,000, capturing approximately 87% of the solid tumor MRD market share. Signatera is expected to achieve around $1.5 billion in revenue in its fifth year, about twice that of Cologuard during the same period. The market consensus opportunity is around $20 billion, with monitoring of cancer survivors, advanced treatment monitoring, CAR-T, and global expansion potentially further scaling up.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.