8 of the best startups in San Francisco with funding, stage and one real downside each, plus 657 open SF startup roles we're tracking on Clera.
The best startups in San Francisco to join right now, if you want a company with fresh funding and real revenue: Harvey for legal AI, Sierra and Decagon for AI agents, Perplexity for consumer AI search, Vercel for developer tools, Mercury for fintech, Ambience Healthcare for health AI, and NewLimit if you want deep science. Best overall for a big, fast step up: Harvey. Best for builders who love developers: Vercel. Best if you want an early lab feel with a big bank account: NewLimit.
One more thing before the list. On Clera this week, 657 open startup roles are based in San Francisco, at 320 companies, and 68% of them are fully in the office. If you were hoping SF startups had gone remote, they mostly have not.
The list at a glance
| Startup | Best for | Key stat | One limitation |
|---|---|---|---|
| Harvey | Legal AI at scale | $11B valuation, March 2026 | Priced for near-perfect execution |
| Sierra | Enterprise AI agents | $950M Series E at $15.8B, 2026 | Long enterprise sales cycles |
| Decagon | Earlier-stage AI agents | $250M Series D at $4.5B, Jan 2026 | Head-to-head with Sierra |
| Perplexity | Consumer AI search | ~$20B valuation, Sept 2025 | IPO target is 2028 |
| Vercel | Developer tools | $300M Series F at $9.3B | Equity priced after a big step-up |
| Mercury | Fintech for startups | $300M Series C at $3.5B | Regulated business, slower shipping |
| Ambience Healthcare | Health AI | $243M Series C, $1.25B valuation | Hospital sales take a while |
| NewLimit | Biotech and longevity | $435M round, ~$3.1B valuation, 2026 | Clinical-stage science risk |
How we picked
We started with roughly 30 San Francisco Bay Area startups that show up across public rankings (Seedtable, Y Combinator's company directory, Built In's 2026 Best Startup Places to Work) and in Clera's own job data. A company made the cut only if it raised a priced round in the last 15 months, has public revenue or customer numbers, and is headquartered in the Bay Area. We dropped companies whose latest round we could not confirm with a news source, and companies in the middle of an acquisition. We also skipped the giants (OpenAI, Anthropic, Databricks), because at that size you are joining a big company, whatever the cap table says. Every number below links to where we found it.
1. Harvey
Best for: people who want a fast-growing AI company with a clear, paying customer base.
Harvey builds AI for lawyers and law firms. In March 2026 it raised $200 million at an $11 billion valuation, co-led by GIC and Sequoia, three months after a round at $8 billion. The company says the money goes into AI agents and the legal engineering teams it embeds with customers, which means hiring on both the product and the customer side.
Why it made the list: real demand from a conservative industry, and a role mix that goes past pure engineering.
Real downside: Techleap's coverage of the round puts the price at roughly 58 times annual revenue. Any slowdown hits equity value hard.
Stage: late-stage, private.
2. Sierra
Best for: engineers and go-to-market people who want to build AI agents for big companies.
Sierra was co-founded by Bret Taylor and Clay Bavor and sells AI agents that handle customer conversations. In 2026 it closed a $950 million Series E at a $15.8 billion valuation, led by Tiger Global and GV, with ARR reported above $150 million. That is up from a $4.5 billion valuation in October 2024.
Why it made the list: one of the fastest valuation climbs in SF, and a product category that is still being defined.
Real downside: enterprise deals are slow and political. Taylor himself has said he expects a market correction within two years.
Stage: late-stage, private.
3. Decagon
Best for: people who want the AI agent space a size or two earlier than Sierra.
Decagon was founded in 2023 by Jesse Zhang and Ashwin Sreenivas and builds AI agents for customer support across chat, email, voice and SMS. In January 2026 it raised a $250 million Series D at a $4.5 billion valuation, triple its previous price. Sacra estimates about $35 million in annualized revenue as of October 2025.
Why it made the list: younger team, so your scope per person is bigger.
Real downside: it sells into the same budget as Sierra and plenty of others. Expect a knife fight on every deal.
Stage: growth-stage.
4. Perplexity
Best for: product people and engineers who want to ship consumer AI used by millions.
Perplexity runs an AI answer engine. It was valued at about $20 billion in September 2025, and in August 2026 Nvidia was reported to be in talks to join a round above $30 billion. The same reporting puts annualized revenue above $750 million.
Why it made the list: consumer scale at a company still small enough that your work is visible.
Real downside: the CEO has said the IPO target is 2028, so equity stays illiquid for a while, and the competition is Google and OpenAI.
Stage: late-stage, private.
5. Vercel
Best for: engineers who care about developer experience and open source.
Vercel makes the cloud platform behind Next.js. In September 2025 it closed a $300 million Series F at a $9.3 billion valuation, co-led by Accel and GIC, and reported 82% year-over-year revenue growth. It also ran a roughly $300 million employee tender, which is a good sign for liquidity.
Why it made the list: a public, much-loved product, plus a track record of letting employees sell shares.
Real downside: the valuation nearly tripled from $3.25 billion in May 2024, so new equity grants are priced high.
Stage: late-stage, private.
6. Mercury
Best for: people who want fintech with startups as the customer.
Mercury offers banking for startups and small businesses. In March 2025 it raised $300 million at a $3.5 billion valuation, led by Sequoia. Wikipedia puts 2024 revenue around $500 million.
Why it made the list: real revenue, a clear customer, and a calmer valuation than most AI names on this list.
Real downside: money products are regulated. Shipping is slower than at a pure software startup.
Stage: growth-stage.
7. Ambience Healthcare
Best for: people who want AI work that ends up in a hospital, not a dashboard.
Ambience Healthcare builds AI for clinical documentation and medical coding. In July 2025 it announced a $243 million Series C, co-led by Oak HC/FT and Andreessen Horowitz, at a $1.25 billion valuation.
Why it made the list: the smallest company here by valuation, so earlier equity and more room to own a whole area.
Real downside: hospitals buy slowly and care a lot about compliance. Expect long pilots.
Stage: growth-stage.
8. NewLimit
Best for: scientists and ML people who want to work on aging biology.
NewLimit develops epigenetic therapies meant to restore younger cell function. In June 2026 it announced $435 million in new funding, led by Founders Fund, at a valuation of about $3.1 billion, to run its first clinical trial of a liver medicine.
Why it made the list: few startups pair this much cash with this early a team.
Real downside: there is no product revenue yet. If the trial disappoints, so does everything else.
Stage: clinical-stage.
What Clera's data says about SF startup hiring
We pulled every open startup role on Clera with a San Francisco location on October 23, 2026:
- 657 open roles at 320 companies. That is about 40% of all open roles on Clera.
- 400 of them are engineering, research or data roles. 89 are sales, growth or marketing.
- 135 have "founding" in the title. If you want to be early, SF is where the early seats are.
- 450 are fully in the office, 103 hybrid, 46 remote. Plan to commute.
- The median posted range is $165,000 to $240,000, across the 571 roles that list a salary.
Nobody else publishes these numbers, so here is the takeaway: SF startups are hiring a lot, mostly engineers, mostly in person, and they pay for it. For a deeper look at pay, see our guide to software engineer salaries at startups.
How to choose
- If you want the biggest step up on your resume, pick Harvey or Sierra. Late-stage, fast growth, known names.
- If you want more scope and earlier equity, pick Decagon or Ambience Healthcare.
- If you care about liquidity, ask every company when it last ran a tender offer. Vercel has a recent one on record.
- If you want to work on science, not software, NewLimit is the rare option with both cash and a clear next milestone.
Whatever you pick, bring the right questions to the interview. Our list of 5 questions to ask before joining an AI startup is a good place to start.
Frequently Asked Questions
Yes. On Clera alone, 657 startup roles are open in San Francisco this week, at 320 companies. Most are engineering roles, and most are in the office. AI companies make up a big share, which matches what we see in our list of AI companies hiring right now.
Mostly not. Of the 657 SF roles on Clera, 450 are fully in the office and only 46 are remote. If remote is a must, look at our list of remote startup jobs instead, or ask about hybrid early in the process.
The median posted range on Clera's SF roles is $165,000 to $240,000 a year, based on 571 roles that list salary. Late-stage companies like the ones above often add equity. Before you sign, read our guide to negotiating your first salary at an AI startup.
Safer for your paycheck, usually. Riskier for your equity upside, because the price is already high. An $11 billion company has to grow a lot for your shares to multiply, while a seed-stage company can multiply many times or go to zero. Pick based on which risk you can live with.
New York is the other big hub. We ranked the best startups to work for in NYC last week, using the same method as this list.
Sources
- Bloomberg Law: Harvey raises at $11 billion valuation
- Trending Topics: Sierra raises $950 million
- Bloomberg: Decagon valued at $4.5 billion
- NewsBytes: Nvidia in talks to invest in Perplexity
- Business Wire: Vercel closes Series F
- Axios: Mercury raises $300M Series C
- Ambience Healthcare: Series C announcement
- NewLimit announces $435 million financing(https://statnews.com/2026/06/02/longevity-startup-newlimit-announces-435-million-clinical-trial-financing/)
- Clera job data, open San Francisco roles as of October 23, 2026
Get SF startups to come to you
Applying to eight companies one by one takes weeks. On Clera you join once, and VC-backed startups bid on you, with a warm intro when there's a fit. 320 companies are hiring in San Francisco on Clera right now. Let them do the chasing.
