Instinct Series C funding closed at $1 billion on 28 September 2026, with Sequoia Capital, Benchmark and Coatue backing the viral personal AI agent at a $10 billion valuation. It is the startup’s fourth priced round in roughly five months, and the second in 33 days. A month earlier the same company was worth $2.5 billion.
Instinct confirmed the round in a Business Wire press release on Monday morning, and TechCrunch’s Sarah Perez reported that the company declined to make founder Noah Shinn available for interviews. What the announcement does not contain matters as much as what it does: no user count, no revenue, no pricing and no date for opening the invite-only service to everyone.
This article sets out what the Instinct Series C announcement says, rebuilds the funding timeline from the public reports, and works through the arithmetic behind a 200-fold markup since the seed round. It then looks at what the money is likely to buy, how Instinct compares with Meta’s Muse, the privacy questions that followed the product from its launch, and what the round means for businesses that are about to start taking calls and bookings from AI agents.
Table of contents
- What the Instinct Series C Announcement Says
- From $50 Million to $10 Billion in About Five Months
- The Arithmetic Inside the Instinct Series C
- Why Investors Backed the Instinct Series C Now: Compute and Demand
- What Instinct Actually Does
- The Founder Behind the Instinct Series C
- Muse, Siri and the Crowded Personal Agent Race
- The Privacy Questions the Instinct Series C Leaves Open
- How Instinct Could Make Money
- What the Instinct Series C Means for Businesses and Users
- Instinct Series C FAQs
- References
What the Instinct Series C Announcement Says
The facts in the release are short. Instinct “has raised an additional $1 billion in Series C funding from Sequoia Capital, Benchmark Capital and Coatue at a $10 billion valuation”, and says the money will help it “bring useful AI to everyone”. The operating company is Spear Street Technology, Inc., based in San Francisco.
The Instinct Series C investors and price
Benchmark is the only name that appears in both of the last two rounds: it co-led the $250 million Series B in August with Index Ventures, and it returns for the Instinct Series C alongside Sequoia and Coatue. The release does not say who led, whether the $10 billion figure is pre-money or post-money, or whether any shares changed hands in a secondary sale.
The statement, and what it leaves out
The only quote is attributed to Shinn: “We’re building Instinct to be the best personal agent that can handle the deeply personal nuances of everyday life. This funding helps us bring Instinct to more people and continue building the future of personal AI. It’s an exciting, creative time, and we’re just getting started.” TechCrunch noted that the startup “has also not shared its user numbers or any growth metrics”.
Where the news came first
The Instinct Series C was not a surprise. On 16 September The Information reported that Spear Street was in talks to raise $1 billion at about $10 billion, with Sequoia and Benchmark discussing a lead and Coatue also in talks. That report also said the user base had grown past 100,000, according to one person, and that the service was straining under the load. The Instinct Series C closed 12 days later on the terms first reported.
From $50 Million to $10 Billion in About Five Months
Instinct’s funding history has been reported in pieces across The Wall Street Journal, Forbes, TechCrunch and The Information. Put together, it shows a company that did not exist in March and is now valued like a mature software business.
The four rounds
Spear Street Technology was registered in California in April 2026, according to Tech Funding News. An early round from Conviction’s Pranav Reddy and Greenoaks’ Neil Mehta valued it at $50 million. Kleiner Perkins partner Mamoon Hamid then led a $75 million Series A at $500 million in early August. About three weeks later, Index Ventures and Benchmark co-led the $250 million Series B, reported as $2.25 billion before the new money and $2.5 billion after it.
| Round | Reported timing | Amount | Valuation | Named investors |
|---|---|---|---|---|
| Early round | Spring 2026 | About $25M (implied) | $50M | Conviction, Greenoaks |
| Series A | Early August 2026 | $75M | $500M | Kleiner Perkins (lead) |
| Series B | 26 August 2026 | $250M | $2.5B post-money | Index Ventures, Benchmark (co-leads) |
| Series C | 28 September 2026 | $1B | $10B | Sequoia, Benchmark, Coatue |
The early round’s size is not published. The figure above is simply the reported $350 million total after the Series B, minus $250 million and $75 million.
Step-ups are shrinking as rounds grow
Each round has multiplied the valuation, but by less each time. The Series A priced the company at 10 times the early round, the Series B at five times the Series A, and the Instinct Series C at four times the Series B. From $50 million to $10 billion is a 200-fold increase.
How fast the rounds arrived
The gaps are as unusual as the multiples. Roughly three weeks separated the Series A from the Series B, and 33 days separated the Series B from the Instinct Series C. Counting from the April registration, the company reached a $10 billion price in about 180 days, while its product was still invite-only.
The Arithmetic Inside the Instinct Series C
Four numbers explain most of what this round implies. None of them requires information Instinct has not published, but each depends on a stated assumption.
One round, nearly three times all earlier capital
Before Monday, Instinct had raised about $350 million in total. The Instinct Series C adds $1 billion on its own, 2.86 times everything that came before, taking the total to about $1.35 billion. A company that raised its first priced round in the spring now holds more cash than most software firms raise in a decade.
Ten per cent of the company, if the figure is post-money
If $10 billion is the post-money valuation, the new investors bought 10% of Instinct. If it is pre-money, they bought about 9.1% of an $11 billion company. The Series B worked the same way: $250 million against $2.5 billion post-money is also 10%. Consistent 10% slices at rising prices are a common way for founders to limit dilution while the price climbs.
$100,000 of valuation per reported user
The only usage figure in the public record is the “more than 100,000” users The Information cited on 16 September. Dividing $10 billion by 100,000 gives $100,000 of valuation per user, and $10,000 of fresh cash per user from the Instinct Series C alone. Those ratios will fall quickly if the waitlist opens, which is the point: investors are paying for the users Instinct does not yet have.
What the Instinct Series C price assumes
A $10 billion valuation for a free product with no disclosed revenue is a bet on three things happening together. Instinct must scale from an invite-only beta to a mass product, keep users once Muse and others offer similar features, and find a business model that people accept. The rest of this article looks at the evidence on each.
Why Investors Backed the Instinct Series C Now: Compute and Demand
The clearest explanation for the size of the Instinct Series C is not in the press release. It is in the reports that preceded it.
A service that says it is at full capacity
The Information reported that Instinct has sometimes told users it is running at full capacity and that “responses may be slower”, as people use it to negotiate bills and answer emails. Shinn said in his August launch post that the beta was invite-only “while we’re actively bringing up more compute”. Demand, in other words, is already ahead of supply.
Agents burn more inference than chatbots
A chatbot answers one message and stops. An agent that books a table, checks a calendar and follows up by text runs many model calls, keeps a virtual computer open and waits on third-party websites. Every extra user adds that load. That makes consumer autonomous agents capital-hungry in a way that earlier chat apps were not, and it explains why the Instinct Series C is four times the size of the round before it.
Invite-only as a capacity valve
An invite list lets a company choose how fast demand grows. It also creates scarcity, which one commentator on X credited as part of “the best consumer product launch strategy since Uber”. The release says only that people can visit app.instinct.co/login “to stay up to date for expanded access”. The Instinct Series C money is best read as fuel for turning that valve.
What Instinct Actually Does
Instinct has no new interface. Users text it or call it, and it works through its own phone number and its own computer, using websites and apps the way a person would.
Text it or call it
Shinn’s August post described it plainly: “The interface is simple: there are no new interfaces. You can text or call it. It’s trained to use a phone and a computer in the same way that humans do.” Early users, he wrote, had planned road trips, bought groceries and concert tickets, and cancelled subscriptions. TechCrunch notes Instinct still has no mobile app.
Concierge, trusted people and location
The release lists the features shipped in recent months. Instinct Concierge makes phone calls for high-touch jobs, such as booking a restaurant that takes no online reservations or joining a dentist’s cancellation list. The Trusted Person Network uses an Instinct-to-Instinct protocol so one user’s agent can coordinate plans with a friend’s agent and share PDFs, images and spreadsheets. Location Sharing on iMessage lets it notice when a user arrives somewhere. We covered the calling feature when it launched, in our look at AI agent calling from Instinct and Muse.
The security claims in the release
The release says “isolated sandboxes, short-lived local credentials, and identity-signed tool execution” have been built in “from day one”. It also describes a new “active detective system” that catches “subtle, nuanced hallucinations” before Instinct acts. None of these claims is accompanied by technical documentation, and the release does not say whether any has been independently tested.
The Founder Behind the Instinct Series C
Investors backing a company this young are backing a person as much as a product. Shinn is 23, and his research record is unusually close to what Instinct does.
Reflexion and agents that learn from failure
Shinn is the first author of Reflexion, a 2023 paper that proposed “language agents” which improve through written self-reflection rather than retraining. Its abstract explains that traditional reinforcement learning methods “require extensive training samples and expensive model fine-tuning”, so Reflexion agents instead “verbally reflect on task feedback signals” and store the lesson in memory for the next attempt. The idea of an agent that learns from its own mistakes sits at the centre of the Instinct Series C pitch.
From tau-bench to Sierra to Spear Street
He is also a co-author of tau-bench, a 2024 benchmark from Sierra that tests whether agents can talk to users, call tools and follow business rules consistently across repeated runs. He worked at Sierra, the customer-service agent company, as a research scientist before leaving in 2025. Both papers are about the gap between a demo that works once and an agent that works every time.
Why investors bet on the person
Frontrun, which tracks the X follows of venture investors, says three of the investors it watches followed Shinn before the Series A, when he had 74 followers, and 22 before the Series B was public. That pattern is typical of a founder-led market, where capital chases a small number of researchers who have shipped something people talk about. It also means the Instinct Series C is priced on expectations about future products as much as on the current one.
Muse, Siri and the Crowded Personal Agent Race
The Instinct Series C landed in a market that changed during September. The biggest change came from Meta.
Meta’s free rival
Meta launched Muse on 8 September. TechCrunch reports that it offers many of Instinct’s features plus deep integration with Meta’s apps, such as summarising Instagram DMs or watching Marketplace listings, and that it has been downloaded millions of times. Muse is also testing human-concierge voice calls, according to SiliconANGLE. We have followed its updates, most recently the Muse agent’s dedicated VM tab.
| Factor | Instinct | Meta Muse |
|---|---|---|
| Access | Invite-only early access | Public app, top of US app stores |
| Interface | SMS, iMessage and phone calls; no app | Mobile app with Meta integrations |
| Price to user | Free for now | Free |
| Phone calls | Instinct Concierge | Testing a concierge call service |
| Social data | Only what the user connects | Instagram, Facebook Groups, Marketplace |
| Usage disclosed | None; 100,000+ reported by The Information | Downloaded millions of times, per TechCrunch |
The rest of the field
SiliconANGLE counts “roughly a dozen prominent AI-agent products” competing to become persistent assistants. Close startup rivals include Wajo, Ollie and Poke. The larger field includes agent modes in OpenAI’s ChatGPT, xAI’s Grok and Apple’s Siri, which Apple rebuilt this year on Apple Intelligence.
Distribution is the real contest
Meta and Apple already sit on billions of phones. Instinct’s answer is to need no app at all, working through the text and call channels every phone already has. The Instinct Series C gives it the money to compete on capacity, but not a distribution channel of its own.
The Privacy Questions the Instinct Series C Leaves Open
An agent that can act for you needs access to almost everything: email, messages, calendar, location and payment details. That access is the product, and it is also where the early problems appeared.
Early incidents on record
TechCrunch reported on 24 August that Instinct’s first terms granted a “perpetual and irrevocable” licence to user material, and that early users had problems disconnecting accounts. Moxxie Ventures founder Katie Jacobs Stanton said the agent “sent an email on my behalf without checking with me first”. Forbes reported a user who asked only for available reservations and was booked into a table with a $200 cancellation fee. We examined how the first terms described email in our piece on the Instinct email launch.
Legal documents unchanged since 26 August
Instinct did fix the licence language. Today “perpetual”, “irrevocable” and “screen capture” do not appear in the terms at all. But both the Terms of Service and the Privacy Policy are still dated 26 August 2026. The Instinct Series C closed 33 days later without either document changing, even though Concierge calling, the Trusted Person Network and Location Sharing shipped in between.
What the documents already say
The terms still appoint Instinct “as your agent to enter into agreements, commitments or transactions on your behalf”, and still cap its liability at the greater of $100 or what the user paid in the previous six months. For a free product, that is $100. The privacy policy is more candid, warning of “unintended payments or communications to outside parties” and of third parties “who may include misleading instructions intended to influence autonomous agents”. That second risk is prompt injection, and it grows with every connected inbox.
How Instinct Could Make Money
A $10 billion Instinct Series C price needs a revenue story, and Instinct has not published one.
Free today, by choice
Instinct is free to use. The Information reported that Shinn has said he does not want to charge users, which suggests the company could try advertising or other models. Forbes reported that a paid subscription tier and an advertising approach drawing on the app’s view of users’ communications and finances were both under consideration.
| Possible model | How it would work | What would have to change |
|---|---|---|
| Subscription | Monthly fee for the agent or for Concierge | Shinn’s stated reluctance to charge users |
| Advertising | Recommendations shaped by paying merchants | A privacy policy that today mentions ads only once, to rule some out |
| Commerce fees | A share of bookings and purchases the agent completes | Merchant agreements, and disclosure to users |
Why the privacy policy matters for any ad model
The current privacy policy uses the word “advertising” in one place: to promise that data from Google Workspace APIs is never used “for serving ads, including retargeting, personalized or interest-based advertising”. It says Instinct does not sell user information. Any business built on ads would need a new policy, and users would be entitled to read it before the agent that holds their inbox starts earning money from it.
What the Instinct Series C Means for Businesses and Users
For most readers the question is not whether $10 billion is the right price for the Instinct Series C. It is what changes when well-funded agents start acting on people’s behalf at scale.
For businesses that take bookings and calls
Restaurants, clinics, salons and service firms should expect more calls and bookings placed by AI agents rather than people. Decide now how staff should handle them. Confirm bookings back to the customer’s own number, make cancellation fees explicit before accepting, and log agent-originated requests. Our overview of AI employees and autonomous AI agents covers how the same technology works inside a business.
For people weighing an invite
Connect only the accounts the agent needs, and read what it can do with them. Turn on the training opt-out the privacy policy offers, check actions before they complete where the product allows, and treat any email the agent reads as something a stranger could try to instruct. The Instinct Series C funds more capacity; it does not change what the terms say.
For founders building agents
The Instinct Series C shows how much capital is available for consumer agents with visible demand, even before revenue. It also shows the cost structure: compute scales with usage, and trust problems surface in public. An AI strategy that budgets for inference, security review and a clear consent model will age better than one built only around a growth curve.
Instinct Series C FAQs
How much has Instinct raised in total?
About $1.35 billion. Reports put the total at $350 million after the August Series B, and the Instinct Series C adds $1 billion.
Who invested in the Instinct Series C?
Sequoia Capital, Benchmark and Coatue. Benchmark also co-led the Series B with Index Ventures. The release did not name a lead investor for the new round.
Is Instinct available to everyone?
No. It remains in invite-only early access, and the company has given no date for opening it up.
Does the Instinct Series C change what users pay?
Not yet. Instinct is free today. Shinn has said he does not want to charge users, and reports say both subscriptions and advertising have been considered.
How does Instinct compare with Meta’s Muse?
Both are free personal agents that act for users. Muse is a public app tied into Meta’s services; Instinct works over text and calls and remains invite-only.
References
Viral AI agent Instinct raises $1B Series C at a $10B valuation (TechCrunch)
Everyday personal AI assistant startup Instinct raises $1B at $10B valuation (SiliconANGLE)
Instinct AI Assistant Targets $10 Billion Valuation (PYMNTS)
Instinct raises $250M Series B at $2.5B (Tech Funding News)
Instinct AI assistant raises $250 million Series B at $2.5B valuation (Quartz via Yahoo Finance)
Reflexion: Language Agents with Verbal Reinforcement Learning (arXiv)
tau-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains (arXiv)
More AI coverage: explore Progressive Robot's AI Models, Tools & Releases hub — hands-on reviews, setup guides and benchmarks in one place.
