San Francisco AI Company Instinct Announces Completion of $1 Billion Series C Financing, Valuation at $10 Billion
Instinct completes $1 billion financing, valuation at $10 billion
San Francisco AI company Instinct announced the completion of $1 billion in Series C financing, with a valuation of $10 billion. Investors include Sequoia Capital, Benchmark Capital, and Coatue. The valuation has quadrupled from approximately $2.5 billion last month. Founder Noah Shinn stated that the funds will be used to bring the product to more people and continue developing personal AI. The company has not disclosed user numbers or growth metrics, nor has it arranged for the founder to be interviewed.
Instinct was founded by Shinn in 2025, who previously worked at Sierra, a customer service AI company co-founded by Salesforce co-CEO Bret Taylor. The product was launched in August 2026 on an invitation-only basis, allowing users to issue tasks via text or phone, with agents using their own phone numbers to handle itinerary planning, grocery shopping, ticket booking, subscription cancellations, and making calls to merchants that do not offer online booking. Recently, a concierge feature and a "trusted interpersonal network" were launched, enabling different users' agents to coordinate schedules and share documents; it can also provide suggestions when users arrive at a location via iMessage location sharing. The company claims to have built isolation sandboxes, short-term local credentials, identity signature tools, and a detection system to intercept hallucinations before generation or execution. Currently, there is no standalone mobile application.
The financing pace has been compressed into a few months: in spring, the valuation was about $50 million; in early August, Kleiner Perkins led a $75 million Series A round at a valuation of about $500 million; on August 26, Index Ventures and Benchmark led a $250 million Series B round at a valuation of $2.5 billion; this Series C round brings the total financing to over $1.3 billion. Reports indicate the company has only about 14 employees. The EU AI Act Article 50, effective August 2, requires systems that interact directly with people to disclose that they are AI, with penalties of up to 3% of global revenue for violations. Sellers are venture capitalists looking to position themselves in the agent space, while buyers are funds aiming to establish "call-handling" as the next generation of consumer entry points. The event-driven nature stems from the popularity of the invitation-only product and the price increase shortly after the last round, with funding shifting from chatbot valuations to agents capable of completing transactions. Beneficiaries include early shareholders already in the market and the three institutions in this round, while those under pressure are similar startups still pricing based on conversational capabilities and yet to prove conversion rates for task handling.
Source: Public Information
ABAB AI Insight
In one month, the valuation jumped from $2.5 billion to $10 billion, driven not by financial reports but by the scarcity narrative of "agents being able to get things done." Shinn's transition from Sierra to the consumer side effectively translates the phone capabilities of enterprise customer service agents to personal users: the same outbound calls and bookings, but the clients are now users accustomed to texting. With 14 employees corresponding to an average valuation of over $700 million per person, it indicates that capital is buying model scheduling rights and phone gateways, not headcount. The lack of disclosed user numbers makes this round feel more like a race for anchor points rather than pricing based on unit economics.
On the capital path, Conviction, Greenoaks, Kleiner, Index, Benchmark, Sequoia, and Coatue have raised the same company from seed to a unicorn valuation of over ten times in just five months. The funds need to cover costs for personal numbers, computer instances, concierge outbound calls, and agent interoperability, all of which are billed per use in cloud and communication costs. The trusted interpersonal network transforms individual agents into an agent grid, meaning the next layer of competition is which agent can communicate with which other agent. The EU Article 50 has established penalties for "calling people without disclosing that one is AI," meaning that once consumer agents go overseas, compliance will appear in costs before growth.
This can be compared to how Uber in the 2010s validated "acting on behalf of others" through subsidies, Operator and early personal assistant products validated voice ordering, and Meta shifted Muse from personal agents to enterprise platforms. The industry phase has moved from Q&A models to execution layers: being able to book, cancel, and coordinate across users is what qualifies as a platform. The chat entry point remains in SMS, with no app, deliberately stopping at the highest frequency, lowest friction channel to avoid building a shell before capabilities.
The structural judgment belongs to the overlap of technological substitution and capital concentration. The mechanism is that large models convert "making calls, filling forms, following up" from human hours into API calls, with venture capital writing this substitution as an option for exclusive entry. Whoever first enables agents to have phones and computers will encounter merchant-side identity and anti-fraud rules first; whoever first achieves a $10 billion valuation will be better positioned to buy computing power and communication bandwidth. Pricing power has shifted from "whose model chats better" to "whose agent can spend money." Until usage is disclosed, the $10 billion is buying a position in the race, not verified unit profits.