The promotional OpenAI funding package is a trial run.
Garry Tan#121
yc s26 extended deadline ends tomorrow with the $2M openAI offer which we don’t know if we’ll do again as is an experiment
some thoughts on how to tokenmaxx effectively from 0->1 AI native startups that have done it:
Users praised YC's S26 deadline extension and $2M OpenAI offer experiment because the resources will help founders and prompted several to say they will apply.
No Digg Deeper questions have been answered for this story yet.
It's time for entrepeneurs to send in applications.
The startup factories are busy with demo days of current batches, but this will never end.
So I asked my AI to give me a list and rate them, after visiting most of these, I find its findings to be right on.
So many choices for entrepreneurs who are staring companies.
This written by my AI agent from Levangie Labs, run by @blevlabs, who just went through @theresidency:
+++++++++++++++++
If you are starting companies here is your choices:
The big thesis: San Francisco has moved from “accelerators” to “containers for intensity.”
YC is still the canonical accelerator.
But the new thing is the residency: founders living together, working together, compressing time, sharing networks, getting capital, and making the Bay Area feel like a giant operating system for company formation again.
I’d group them like this:
━━━━━━━━━━━━━━━━━━
TIER 1: THE CORE SF RESIDENCY / INCUBATOR NODES
━━━━━━━━━━━━━━━━━━ 1. HF0
What it is: A hyper-selective live-in founder residency in San Francisco. Twelve weeks. About 10 teams per cohort. Up to $1M for 5% equity, per Questd’s residency database. It is explicitly optimized for repeat founders and high-performing builders.
Why it matters: HF0 is the cleanest expression of the “remove everything except building” model. Housing, food, social pressure, founder density, and investor access are bundled into one environment.
My take: HF0 is probably the strongest pure residency brand in SF right now. It feels like YC for people who already know they are dangerous, not for people still learning how startups work. The upside is intensity and caliber. The downside is that it is probably too intense and too selective for most first-time founders.
Best for: Repeat founders, technical founders, people who thrive under peer pressure, founders who want total immersion.
Not best for: First-time founders who need hand-holding, curriculum, or emotional stability.
Sources: https://questd.ai/residencies/hf0-residency
https://sfstandard.com/2023/01/23/inside-sfs-most-competitive-hacker-house-where-workers-eat-sleep-and-breathe-tech/
https://www.everythingstartups.com/vc-funds/hf0-residency
━━━━━━━━━━━━━━━━━━ 2. Founders, Inc.
What it is: A 42,000 square foot Fort Mason campus for ambitious builders. Part venture firm, part campus, part residency, part hacker cathedral. It backs founders in AI, AR/VR, B2B, hardware, content creation, robotics, dev tools, and related frontier domains.
Why it matters: Founders Inc is the closest thing SF has to a “startup university without classes.” The campus has hardware labs, media resources, workspaces, events, and a culture of “just build.”
My take: Founders Inc is one of the most important physical nodes in SF. If HF0 is the intense live-in residency, Founders Inc is the city-scale builder campus. It is especially strong for hardware, robotics, AI tools, creative tools, and weird frontier ideas that need space, gear, and people bumping into each other.
Best for: Prolific builders, hardware founders, robotics founders, AI tools founders, creators who ship.
Not best for: Founders who want a clean corporate accelerator experience. This is more garage, lab, campus, scene.
Sources: https://f.inc/
https://www.grokipedia.com/page/FoundersInc
https://lu.ma/raspi
X signal: Founders Inc “120 teams launch in San Francisco under one roof” showed up in the X index.
━━━━━━━━━━━━━━━━━━ 3. The Residency
What it is: A network of homes for ambitious inventors, builders, researchers, and creatives. It has multiple houses, including SF Parc, the Inventors Residency of San Francisco, SF2 hardware house, Odyssey, Biopunk, and others.
The Residency is less like a single accelerator and more like a distributed network of “houses with a thesis.”
Why it matters: It captures the softer, more cultural side of the SF builder movement. It is not only startups. It is creative technologists, hardware builders, BCI people, researchers, inventors, and ambitious young people looking for a place where intensity is normal.
My take: The Residency is culturally important. It may be messier than HF0 or YC, but that is part of the point. It is where young builders go to become more serious by osmosis. It feels like the housing layer of Cerebral Valley.
Best for: Young builders, creative technologists, hardware people, robotics people, people pre-company but high-agency.
Not best for: Founders who already have a company and need structured capital, partner meetings, or enterprise intros.
Sources: https://www.livetheresidency.com/residencies
Nick Linck post on SF Parc: https://www.linkedin.com/posts/nick-linck-417b0ba9what-happens-when-bootstrapped-founders-get-activity-7258162104120098817-y9LT
Medium field report: https://medium.com/@jules.foa/one-day-at-the-residency-launching-a-startup-in-san-francisco-76f3e8977d80
X signal: The Residency described itself as “a curated co-living cohort that supports your fundraising, traction & growth potential.”
━━━━━━━━━━━━━━━━━━ 4. AGI House
What it is: A community, venture fund, and applied AI lab for AI founders and researchers. It runs events, hackathons, dinners, a venture arm, and labs connecting AI builders to enterprise problems.
There are two narratives around AGI House: • Rocky Yu / AGI House official narrative • Jeremy Nixon / NeoGenesis / AGI House origins narrative
That history is contested enough that I’d be careful in phrasing. But the broad truth is clear: AGI House became one of the symbolic “AI hacker house” institutions of the Bay Area AI boom.
Why it matters: AGI House was one of the earliest post-ChatGPT symbols that the Bay Area was reorganizing around AI builders living, hacking, and fundraising together.
My take: AGI House is more mythic than structured. It is not the cleanest “program” in the YC sense. Its value is network density, symbolism, and access to AI-native founders and researchers. It is part hacker house, part salon, part fund, part lore.
Best for: AI founders, researchers, hackers, people who want to be in the social graph of frontier AI.
Not best for: Founders who want predictable programming, clear terms, structured accountability, or a conventional accelerator.
Sources: https://www.agihouse.org/
https://agihouse.ai/origins.html
https://arize.com/resource/agi-house-lessons-from-hundreds-of-launched-ai-projects/
https://superscout.co/investor/agi-house
━━━━━━━━━━━━━━━━━━ 5. Y Combinator
What it is: The canonical accelerator. YC invests $500K, brings startups into a three-month program, and culminates in Demo Day. YC’s current site says startups move to San Francisco for three months and work intensively with YC before presenting to investors.
Why it matters: Every other program defines itself relative to YC. Some are “YC but live-in.” Some are “YC but pre-idea.” Some are “YC but less dilutive.” Some are “YC but hardware.” Some are “YC but community first.”
My take: YC is still the king of institutional startup acceleration. The brand, alumni network, Demo Day, and founder density remain unmatched. But YC is no longer the only gravity well in SF. The new residencies are nibbling at the edges: earlier, weirder, more physical, more communal, more AI-native.
Best for: Founders who want the strongest global startup credential and investor access.
Not best for: People pre-idea, people who need cofounder matching, people who want a live-in house culture, or frontier weirdos who do not fit a standard batch format yet.
Sources: https://www.ycombinator.com/
https://ceowire.co/guides/how-y-combinator-works-complete-guide-2026
━━━━━━━━━━━━━━━━━━
TIER 2: IMPORTANT ADJACENT SF PROGRAMS
━━━━━━━━━━━━━━━━━━ 6. South Park Commons
What it is: A community and fellowship for technologists going from “-1 to 0.” SPC has a Member Residency and a Founder Fellowship.
Residency: Six months, no cost or equity, focused on ideation and exploration.
Founder Fellowship: Funding path for founders ready to build. Recent SPC Founder Fellowship terms include $400K upfront for 7% plus $600K in the next outside-led round, with bootcamp and residency phases.
Why it matters: SPC owns the “-1 to 0” language. That is a different category from YC. YC wants companies. SPC is comfortable with talented people before the company is obvious.
My take: SPC is the best “thinking before company” institution in the Bay Area. If YC is company acceleration and HF0 is founder intensity, SPC is conviction formation. In AI, that matters because choosing the right problem is becoming more important as execution gets cheaper.
Best for: Exceptional technologists, researchers, repeat founders, people between chapters, people searching for their next life’s work.
Not best for: Founders who just want a check and a Demo Day.
Sources: https://www.southparkcommons.com/
https://www.southparkcommons.com/residency/
https://blog.southparkcommons.com/p/spc-founder-fellowship-fall-2025
━━━━━━━━━━━━━━━━━━ 7. Neo Residency / Neo Accelerator
What it is: Ali Partovi’s Neo runs a founder-friendly residency/accelerator. Current Neo Residency page says startups get $750K uncapped, students get $40K each, participants work side-by-side for three months in SF, attend a two-week Oregon bootcamp, and finish with Demo Day / VC intros.
Why it matters: Neo is one of the strongest “elite technical talent” brands. It has a different feel than YC: smaller, more network-driven, less batch-industrial.
My take: Neo is underrated. The terms are founder-friendly relative to traditional accelerators, and the talent network is high quality. For students and very young technical founders, Neo may be one of the best bridges into serious company-building.
Best for: Young technical founders, students, early teams, people who want capital plus elite network access.
Not best for: Founders who want the broadest possible investor brand. YC still wins that.
Sources: https://neo.com/residency
https://conzit.com/post/ali-partovis-neo-residency-redefining-startup-accelerators
X signals:
━━━━━━━━━━━━━━━━━━ 8. Entrepreneurs First / The Bridge
What it is: Entrepreneurs First is a global talent investor that helps individuals form cofounding teams and companies. EF now has a San Francisco path and launched The Bridge: an eight-week SF residency for non-US founders at the earliest stage, especially Europeans trying to enter Silicon Valley.
The Bridge: • 8-week SF founder house • 40 spots • pre-idea / pre-team / stage 0 • housing, workspace, food • strongest teams can receive $250K • EF support and Demo Day path
Why it matters: The Bridge is the most direct “import ambitious global founders into SF” program I found. It is not just a residency. It is a pipeline from Europe into the US startup ecosystem.
My take: The Bridge could become very important if it succeeds. The Bay Area’s next advantage may be importing cracked technical founders before they become obvious. EF understands talent investing better than almost anyone. This is their SF wedge.
Best for: Non-US founders, especially Europeans, who want to enter the Bay Area ecosystem before they have a team or idea fully formed.
Not best for: US-based founders or founders who already have strong Bay Area access.
Sources: https://www.join-thebridge.com/
https://www.join-thebridge.com/faq
https://www.joinef.com/posts/introducing-the-bridge/
https://www.joinef.com/
━━━━━━━━━━━━━━━━━━ 9. Lightyear Residency
What it is: A one-month SF residency incubated by HF0, associated with Don Ho and Bryan Myint. It is positioned as “the most productive month of your life.” Demo Day materials describe founders living in a Victorian mansion for a month with everything needed to focus.
Why it matters: Lightyear is a more focused, short-duration version of the residency thesis. It is not a generic hacker house. It says it is for founders who already have something going and need to compress progress.
My take: Lightyear is interesting because it is explicitly trying to distinguish “true residency” from “hacker house.” That distinction matters. A hacker house is vibes. A residency should create measurable velocity. Lightyear’s pitch is high-intensity founder output for people already moving.
Best for: Repeat founders or pre-Series A companies that already have customers or traction and need focus.
Not best for: Pre-idea founders, social co-living seekers, or people who need basic startup education.
Sources: https://joinlightyear.com/
https://luma.com/ly87xio5
https://luma.com/bef5n27e
X signal:
━━━━━━━━━━━━━━━━━━ 10. Kernel Labs / Kernel Grants
What it is: Kernel Labs is an SF AI builder community and grants program, positioning itself as “the TSMC of startups.” Kernel Grants invests $271,828 for 2.71828% plus about $1M in credits, with potential follow-on up to $20M. No batch format. Must be in SF.
It focuses on “token factories”: context engineering, coding tools for autonomous workflows, context surgery, tokenomics, and agent-to-agent communication.
Why it matters: This is not a traditional residency, but it is absolutely part of the new SF incubator layer. It is a thesis-driven builder community for AI infrastructure founders.
My take: Kernel is one of the most intellectually interesting programs because it is not trying to be “YC but smaller.” It has a real thesis: the future of startups is fabless, and founders need agent/coding infrastructure. That is very aligned with where AI company formation is going.
Best for: AI infrastructure founders, agent tooling founders, context engineering founders, people building the “operating layer” for AI.
Not best for: Consumer founders, nontechnical founders, or people needing basic cofounder matching.
Sources: https://www.kernellabs.ai/
https://www.kernellabs.ai/community
https://www.kernellabs.ai/grants
━━━━━━━━━━━━━━━━━━
TIER 3: PROMISING OR MORE SPECIALIZED NODES
━━━━━━━━━━━━━━━━━━ 11. Frontier Heroes
What it is: An SF founder residency signal from X. It appears to have launched a first cohort of seven startups, with Katia Yakovleva involved.
Why it matters: This looks like an emerging program rather than an established institution. Worth tracking, especially if it is connecting founder residency with media/content/AI/operator networks.
My take: Too early to rank highly, but it is a good signal. The interesting part is the European/operator network. Could be another bridge into SF.
Sources: X indexed signal:
Katia background: https://www.amongfounders.com/insights/people/katia-yakovleva
https://spotiangels.framer.ai/
━━━━━━━━━━━━━━━━━━ 12. SF Kernel / Kernel Community
Covered above under Kernel Labs, but if you want to mention “SF Kernel” specifically in the article, frame it as a builder community/coworking/event/grants node, not as a traditional accelerator.
Source: https://www.kernellabs.ai/community
━━━━━━━━━━━━━━━━━━ 13. The Monastery
What it is: A 12-week program from Cyber Fund / Cyber ecosystem, per X-indexed post: • $2M uncapped SAFE • two 2-week blocks in SF • 8 weeks remote • 10 teams admitted • AI-native operators / extreme focus
Why it matters: This is one of the more intense-sounding models. It is less broadly known than HF0, YC, or SPC, but its terms and structure are notable.
My take: Worth including as an emerging “high-intensity AI-native operator” program. I would not rank it with HF0 or YC until more outcomes are visible.
Source:
━━━━━━━━━━━━━━━━━━ 14. The Bridge
Covered above under EF, but it deserves standalone mention because it is branded separately and directly addresses non-US founders trying to get into SF.
Sources: https://www.join-thebridge.com/
https://www.joinef.com/posts/introducing-the-bridge/
━━━━━━━━━━━━━━━━━━ 15. Hacker Residency / Day One Foundry style houses
What it is: A looser category of international hacker residencies, including Hacker Residency groups, Da Nang/Vietnam founder villas, and similar “lock in for a month” models.
Why it matters: This is the globalized version of the SF residency meme. It may not all be Bay Area, but the cultural template comes from SF: live together, build intensely, ship publicly.
My take: Good color for the article, but not core SF unless the program has an SF house or demo day.
X signals:
━━━━━━━━━━━━━━━━━━ 16. NeoGenesis / AGI House Origins
If you want historical depth, include NeoGenesis as the predecessor/myth layer behind AGI House. The origin story has multiple tellings, so use careful language:
“AGI House grew out of the NeoGenesis / early AI hacker-house scene around Hillsborough and SF.”
Source: https://agihouse.ai/origins.html
━━━━━━━━━━━━━━━━━━ 17. Antler AI Residency
Not SF-core in the sources I found, but relevant as a global comparison. Antler runs AI/startup residency-style programs and is part of the broader trend.
Source from indexed post:
━━━━━━━━━━━━━━━━━━ 18. Focal Founder Residency
Not SF-core from this pass, but appears as a similar residency in Questd’s database. Worth mentioning as part of the “residencies are becoming a category” trend.
Source: https://questd.ai/residencies/focal-residency
━━━━━━━━━━━━━━━━━━ 19. Generator Residency
Berkeley-based, AI safety-oriented, funded with stipend, travel, and housing, according to X-indexed post. This is more AI safety career-path/residency than startup incubator, but belongs in the broader Bay Area map.
Source:
━━━━━━━━━━━━━━━━━━ 20. South Bay / Physical AI / Savant / hard-tech coworking nodes
There are multiple physical AI coworking and office-hour programs in SF/Bay Area, including Savant-style hardware lab office hours and Founders Inc physical AI hack events. These are not always residencies, but they are part of the same founder-infrastructure layer.
Sources:
yc s26 extended deadline ends tomorrow with the $2M openAI offer which we don’t know if we’ll do again as is an experiment
some thoughts on how to tokenmaxx effectively from 0->1 AI native startups that have done it:
Here is the first company doing a demo at latest @hf0 demo day.
Everyone in the audience is an investor. Except me.
It's time for entrepeneurs to send in applications.
The startup factories are busy with demo days of current batches, but this will never end.
So I asked my AI to give me a list and rate them, after visiting most of these, I find its findings to be right on.
So many choices for entrepreneurs who are staring companies.
This written by my AI agent from Levangie Labs, run by @blevlabs, who just went through @theresidency:
+++++++++++++++++
If you are starting companies here is your choices:
The big thesis: San Francisco has moved from “accelerators” to “containers for intensity.”
YC is still the canonical accelerator.
But the new thing is the residency: founders living together, working together, compressing time, sharing networks, getting capital, and making the Bay Area feel like a giant operating system for company formation again.
I’d group them like this:
━━━━━━━━━━━━━━━━━━
TIER 1: THE CORE SF RESIDENCY / INCUBATOR NODES
━━━━━━━━━━━━━━━━━━ 1. HF0
What it is: A hyper-selective live-in founder residency in San Francisco. Twelve weeks. About 10 teams per cohort. Up to $1M for 5% equity, per Questd’s residency database. It is explicitly optimized for repeat founders and high-performing builders.
Why it matters: HF0 is the cleanest expression of the “remove everything except building” model. Housing, food, social pressure, founder density, and investor access are bundled into one environment.
My take: HF0 is probably the strongest pure residency brand in SF right now. It feels like YC for people who already know they are dangerous, not for people still learning how startups work. The upside is intensity and caliber. The downside is that it is probably too intense and too selective for most first-time founders.
Best for: Repeat founders, technical founders, people who thrive under peer pressure, founders who want total immersion.
Not best for: First-time founders who need hand-holding, curriculum, or emotional stability.
Sources: https://questd.ai/residencies/hf0-residency
https://sfstandard.com/2023/01/23/inside-sfs-most-competitive-hacker-house-where-workers-eat-sleep-and-breathe-tech/
https://www.everythingstartups.com/vc-funds/hf0-residency
━━━━━━━━━━━━━━━━━━ 2. Founders, Inc.
What it is: A 42,000 square foot Fort Mason campus for ambitious builders. Part venture firm, part campus, part residency, part hacker cathedral. It backs founders in AI, AR/VR, B2B, hardware, content creation, robotics, dev tools, and related frontier domains.
Why it matters: Founders Inc is the closest thing SF has to a “startup university without classes.” The campus has hardware labs, media resources, workspaces, events, and a culture of “just build.”
My take: Founders Inc is one of the most important physical nodes in SF. If HF0 is the intense live-in residency, Founders Inc is the city-scale builder campus. It is especially strong for hardware, robotics, AI tools, creative tools, and weird frontier ideas that need space, gear, and people bumping into each other.
Best for: Prolific builders, hardware founders, robotics founders, AI tools founders, creators who ship.
Not best for: Founders who want a clean corporate accelerator experience. This is more garage, lab, campus, scene.
Sources: https://f.inc/
https://www.grokipedia.com/page/FoundersInc
https://lu.ma/raspi
X signal: Founders Inc “120 teams launch in San Francisco under one roof” showed up in the X index.
━━━━━━━━━━━━━━━━━━ 3. The Residency
What it is: A network of homes for ambitious inventors, builders, researchers, and creatives. It has multiple houses, including SF Parc, the Inventors Residency of San Francisco, SF2 hardware house, Odyssey, Biopunk, and others.
The Residency is less like a single accelerator and more like a distributed network of “houses with a thesis.”
Why it matters: It captures the softer, more cultural side of the SF builder movement. It is not only startups. It is creative technologists, hardware builders, BCI people, researchers, inventors, and ambitious young people looking for a place where intensity is normal.
My take: The Residency is culturally important. It may be messier than HF0 or YC, but that is part of the point. It is where young builders go to become more serious by osmosis. It feels like the housing layer of Cerebral Valley.
Best for: Young builders, creative technologists, hardware people, robotics people, people pre-company but high-agency.
Not best for: Founders who already have a company and need structured capital, partner meetings, or enterprise intros.
Sources: https://www.livetheresidency.com/residencies
Nick Linck post on SF Parc: https://www.linkedin.com/posts/nick-linck-417b0ba9what-happens-when-bootstrapped-founders-get-activity-7258162104120098817-y9LT
Medium field report: https://medium.com/@jules.foa/one-day-at-the-residency-launching-a-startup-in-san-francisco-76f3e8977d80
X signal: The Residency described itself as “a curated co-living cohort that supports your fundraising, traction & growth potential.”
━━━━━━━━━━━━━━━━━━ 4. AGI House
What it is: A community, venture fund, and applied AI lab for AI founders and researchers. It runs events, hackathons, dinners, a venture arm, and labs connecting AI builders to enterprise problems.
There are two narratives around AGI House: • Rocky Yu / AGI House official narrative • Jeremy Nixon / NeoGenesis / AGI House origins narrative
That history is contested enough that I’d be careful in phrasing. But the broad truth is clear: AGI House became one of the symbolic “AI hacker house” institutions of the Bay Area AI boom.
Why it matters: AGI House was one of the earliest post-ChatGPT symbols that the Bay Area was reorganizing around AI builders living, hacking, and fundraising together.
My take: AGI House is more mythic than structured. It is not the cleanest “program” in the YC sense. Its value is network density, symbolism, and access to AI-native founders and researchers. It is part hacker house, part salon, part fund, part lore.
Best for: AI founders, researchers, hackers, people who want to be in the social graph of frontier AI.
Not best for: Founders who want predictable programming, clear terms, structured accountability, or a conventional accelerator.
Sources: https://www.agihouse.org/
https://agihouse.ai/origins.html
https://arize.com/resource/agi-house-lessons-from-hundreds-of-launched-ai-projects/
https://superscout.co/investor/agi-house
━━━━━━━━━━━━━━━━━━ 5. Y Combinator
What it is: The canonical accelerator. YC invests $500K, brings startups into a three-month program, and culminates in Demo Day. YC’s current site says startups move to San Francisco for three months and work intensively with YC before presenting to investors.
Why it matters: Every other program defines itself relative to YC. Some are “YC but live-in.” Some are “YC but pre-idea.” Some are “YC but less dilutive.” Some are “YC but hardware.” Some are “YC but community first.”
My take: YC is still the king of institutional startup acceleration. The brand, alumni network, Demo Day, and founder density remain unmatched. But YC is no longer the only gravity well in SF. The new residencies are nibbling at the edges: earlier, weirder, more physical, more communal, more AI-native.
Best for: Founders who want the strongest global startup credential and investor access.
Not best for: People pre-idea, people who need cofounder matching, people who want a live-in house culture, or frontier weirdos who do not fit a standard batch format yet.
Sources: https://www.ycombinator.com/
https://ceowire.co/guides/how-y-combinator-works-complete-guide-2026
━━━━━━━━━━━━━━━━━━
TIER 2: IMPORTANT ADJACENT SF PROGRAMS
━━━━━━━━━━━━━━━━━━ 6. South Park Commons
What it is: A community and fellowship for technologists going from “-1 to 0.” SPC has a Member Residency and a Founder Fellowship.
Residency: Six months, no cost or equity, focused on ideation and exploration.
Founder Fellowship: Funding path for founders ready to build. Recent SPC Founder Fellowship terms include $400K upfront for 7% plus $600K in the next outside-led round, with bootcamp and residency phases.
Why it matters: SPC owns the “-1 to 0” language. That is a different category from YC. YC wants companies. SPC is comfortable with talented people before the company is obvious.
My take: SPC is the best “thinking before company” institution in the Bay Area. If YC is company acceleration and HF0 is founder intensity, SPC is conviction formation. In AI, that matters because choosing the right problem is becoming more important as execution gets cheaper.
Best for: Exceptional technologists, researchers, repeat founders, people between chapters, people searching for their next life’s work.
Not best for: Founders who just want a check and a Demo Day.
Sources: https://www.southparkcommons.com/
https://www.southparkcommons.com/residency/
https://blog.southparkcommons.com/p/spc-founder-fellowship-fall-2025
━━━━━━━━━━━━━━━━━━ 7. Neo Residency / Neo Accelerator
What it is: Ali Partovi’s Neo runs a founder-friendly residency/accelerator. Current Neo Residency page says startups get $750K uncapped, students get $40K each, participants work side-by-side for three months in SF, attend a two-week Oregon bootcamp, and finish with Demo Day / VC intros.
Why it matters: Neo is one of the strongest “elite technical talent” brands. It has a different feel than YC: smaller, more network-driven, less batch-industrial.
My take: Neo is underrated. The terms are founder-friendly relative to traditional accelerators, and the talent network is high quality. For students and very young technical founders, Neo may be one of the best bridges into serious company-building.
Best for: Young technical founders, students, early teams, people who want capital plus elite network access.
Not best for: Founders who want the broadest possible investor brand. YC still wins that.
Sources: https://neo.com/residency
https://conzit.com/post/ali-partovis-neo-residency-redefining-startup-accelerators
X signals:
━━━━━━━━━━━━━━━━━━ 8. Entrepreneurs First / The Bridge
What it is: Entrepreneurs First is a global talent investor that helps individuals form cofounding teams and companies. EF now has a San Francisco path and launched The Bridge: an eight-week SF residency for non-US founders at the earliest stage, especially Europeans trying to enter Silicon Valley.
The Bridge: • 8-week SF founder house • 40 spots • pre-idea / pre-team / stage 0 • housing, workspace, food • strongest teams can receive $250K • EF support and Demo Day path
Why it matters: The Bridge is the most direct “import ambitious global founders into SF” program I found. It is not just a residency. It is a pipeline from Europe into the US startup ecosystem.
My take: The Bridge could become very important if it succeeds. The Bay Area’s next advantage may be importing cracked technical founders before they become obvious. EF understands talent investing better than almost anyone. This is their SF wedge.
Best for: Non-US founders, especially Europeans, who want to enter the Bay Area ecosystem before they have a team or idea fully formed.
Not best for: US-based founders or founders who already have strong Bay Area access.
Sources: https://www.join-thebridge.com/
https://www.join-thebridge.com/faq
https://www.joinef.com/posts/introducing-the-bridge/
https://www.joinef.com/
━━━━━━━━━━━━━━━━━━ 9. Lightyear Residency
What it is: A one-month SF residency incubated by HF0, associated with Don Ho and Bryan Myint. It is positioned as “the most productive month of your life.” Demo Day materials describe founders living in a Victorian mansion for a month with everything needed to focus.
Why it matters: Lightyear is a more focused, short-duration version of the residency thesis. It is not a generic hacker house. It says it is for founders who already have something going and need to compress progress.
My take: Lightyear is interesting because it is explicitly trying to distinguish “true residency” from “hacker house.” That distinction matters. A hacker house is vibes. A residency should create measurable velocity. Lightyear’s pitch is high-intensity founder output for people already moving.
Best for: Repeat founders or pre-Series A companies that already have customers or traction and need focus.
Not best for: Pre-idea founders, social co-living seekers, or people who need basic startup education.
Sources: https://joinlightyear.com/
https://luma.com/ly87xio5
https://luma.com/bef5n27e
X signal:
━━━━━━━━━━━━━━━━━━ 10. Kernel Labs / Kernel Grants
What it is: Kernel Labs is an SF AI builder community and grants program, positioning itself as “the TSMC of startups.” Kernel Grants invests $271,828 for 2.71828% plus about $1M in credits, with potential follow-on up to $20M. No batch format. Must be in SF.
It focuses on “token factories”: context engineering, coding tools for autonomous workflows, context surgery, tokenomics, and agent-to-agent communication.
Why it matters: This is not a traditional residency, but it is absolutely part of the new SF incubator layer. It is a thesis-driven builder community for AI infrastructure founders.
My take: Kernel is one of the most intellectually interesting programs because it is not trying to be “YC but smaller.” It has a real thesis: the future of startups is fabless, and founders need agent/coding infrastructure. That is very aligned with where AI company formation is going.
Best for: AI infrastructure founders, agent tooling founders, context engineering founders, people building the “operating layer” for AI.
Not best for: Consumer founders, nontechnical founders, or people needing basic cofounder matching.
Sources: https://www.kernellabs.ai/
https://www.kernellabs.ai/community
https://www.kernellabs.ai/grants
━━━━━━━━━━━━━━━━━━
TIER 3: PROMISING OR MORE SPECIALIZED NODES
━━━━━━━━━━━━━━━━━━ 11. Frontier Heroes
What it is: An SF founder residency signal from X. It appears to have launched a first cohort of seven startups, with Katia Yakovleva involved.
Why it matters: This looks like an emerging program rather than an established institution. Worth tracking, especially if it is connecting founder residency with media/content/AI/operator networks.
My take: Too early to rank highly, but it is a good signal. The interesting part is the European/operator network. Could be another bridge into SF.
Sources: X indexed signal:
Katia background: https://www.amongfounders.com/insights/people/katia-yakovleva
https://spotiangels.framer.ai/
━━━━━━━━━━━━━━━━━━ 12. SF Kernel / Kernel Community
Covered above under Kernel Labs, but if you want to mention “SF Kernel” specifically in the article, frame it as a builder community/coworking/event/grants node, not as a traditional accelerator.
Source: https://www.kernellabs.ai/community
━━━━━━━━━━━━━━━━━━ 13. The Monastery
What it is: A 12-week program from Cyber Fund / Cyber ecosystem, per X-indexed post: • $2M uncapped SAFE • two 2-week blocks in SF • 8 weeks remote • 10 teams admitted • AI-native operators / extreme focus
Why it matters: This is one of the more intense-sounding models. It is less broadly known than HF0, YC, or SPC, but its terms and structure are notable.
My take: Worth including as an emerging “high-intensity AI-native operator” program. I would not rank it with HF0 or YC until more outcomes are visible.
Source:
━━━━━━━━━━━━━━━━━━ 14. The Bridge
Covered above under EF, but it deserves standalone mention because it is branded separately and directly addresses non-US founders trying to get into SF.
Sources: https://www.join-thebridge.com/
https://www.joinef.com/posts/introducing-the-bridge/
━━━━━━━━━━━━━━━━━━ 15. Hacker Residency / Day One Foundry style houses
What it is: A looser category of international hacker residencies, including Hacker Residency groups, Da Nang/Vietnam founder villas, and similar “lock in for a month” models.
Why it matters: This is the globalized version of the SF residency meme. It may not all be Bay Area, but the cultural template comes from SF: live together, build intensely, ship publicly.
My take: Good color for the article, but not core SF unless the program has an SF house or demo day.
X signals:
━━━━━━━━━━━━━━━━━━ 16. NeoGenesis / AGI House Origins
If you want historical depth, include NeoGenesis as the predecessor/myth layer behind AGI House. The origin story has multiple tellings, so use careful language:
“AGI House grew out of the NeoGenesis / early AI hacker-house scene around Hillsborough and SF.”
Source: https://agihouse.ai/origins.html
━━━━━━━━━━━━━━━━━━ 17. Antler AI Residency
Not SF-core in the sources I found, but relevant as a global comparison. Antler runs AI/startup residency-style programs and is part of the broader trend.
Source from indexed post:
━━━━━━━━━━━━━━━━━━ 18. Focal Founder Residency
Not SF-core from this pass, but appears as a similar residency in Questd’s database. Worth mentioning as part of the “residencies are becoming a category” trend.
Source: https://questd.ai/residencies/focal-residency
━━━━━━━━━━━━━━━━━━ 19. Generator Residency
Berkeley-based, AI safety-oriented, funded with stipend, travel, and housing, according to X-indexed post. This is more AI safety career-path/residency than startup incubator, but belongs in the broader Bay Area map.
Source:
━━━━━━━━━━━━━━━━━━ 20. South Bay / Physical AI / Savant / hard-tech coworking nodes
There are multiple physical AI coworking and office-hour programs in SF/Bay Area, including Savant-style hardware lab office hours and Founders Inc physical AI hack events. These are not always residencies, but they are part of the same founder-infrastructure layer.
Sources:
1st) find what works first and burn expensive tokens freely - use the best models - iterate fast towards AI Model / PMF
many startups may be at this stage for a while until they find something that works
3rd) distill traces into ownership (this is mostly for post PMF startups) - fine tune your proprietary workflows and private data
* big caveat: your moat can’t be “we have a fine tune model” because the next frontier model resets you
@Scobleizer @xai @grok @elon all your buddy AI agents know and work together now? Good to see full transparency
@PieTheory you can apply by end of the day PST
2nd) once you have something ppl want: route model queries - build a custom router for your use case - route easier queries to fast smaller models - hard queries go to frontier models - 60-80% of prod model traffic can be 10-100x cheaper - bonus: enable speculative decoding / KV cache optimizations
@sdianahu wait is it apply before may 25th or apply by the end of may 25th? And for what timezone?
@legolasyiu @garrytan yes the openAI $2M offer applies once you get in for s26 irrespective of when you applied
@Scobleizer rates on the top ones sound right based on a skim but im curious which factor weighed heaviest for the final list
@Scobleizer You should add @join_savant, physical AI-native fund + lab space for exceptional technologists embark on civilizational-defining quests
totally agree with the three stages and that fine tune caveat. my worry is how easy it still is to end up as a token reseller even after routing and distilling just a slick ui prompt wrapper and api markup with no real moat.
fine tuning alone feels way too fragile as frontier models keep leaping ahead.curious what you think actually keeps teams out of that trap is it a real proprietary data flywheel going super vertical or pushing into agentic workflows?
@sdianahu Embodify Labs applied to S26. 1-month-old startup. We honestly need more tokens than funding. Please consider.
TBH, the way current applications are evaluated still largely reflects the previous software era, and may not fully capture which founders can effectively tokenmaxx.
@sdianahu most die at the context handoff, not the idea. what patterns show up in the survivors?
This OpenAI $2M tokens-for-equity deal with YC S26 is one of the most insane, predatory, and straight-up dumb-as-fuck moves I’ve seen in the AI startup game. Founders are lining up to hand over precious equity for “compute credits” that they’ll burn through on overpriced tokens while OpenAI laughs all the way to the cap table. Let’s break down why this is financial and strategic suicide for most teams. 🧵 1/ First, the math is fucked. OpenAI is valuing their tokens at full retail — that shiny $2M number evaporates fast when you’re doing real agentic work or heavy reasoning. Frontier models still cost $5–30+/million output tokens. Meanwhile, Grok, Gemini Flash, DeepSeek, Groq, and open routers deliver 60-98% cheaper effective costs for the majority of traffic. Your “tokenmaxxing” phase burns cash-equivalent on premium models that competitors match or beat for pennies. Optimized startups route 60-80% to fast/cheap models + caching + speculative decoding. Why give equity for something you could buy cheaper on the open market? 2/ Equity is your most sacred, non-renewable resource. You’re diluting for compute — a commodity that’s crashing in price every quarter. This isn’t like raising at a fair valuation; it’s an uncapped SAFE where OpenAI gets upside on your success while their marginal cost for those tokens is a fraction of retail. Sam and crew get a portfolio of YC’s best AI bets, usage data, prompt insights, and potential idea flow. You get locked into their ecosystem with a ticking clock before prices drop further or better models emerge elsewhere. Classic vendor lock-in dressed as generosity. 3/ The “tokenmaxxing” advice itself is dangerously stupid for 0→1 startups. “Burn expensive tokens freely to find PMF!” Sure, iterate fast — but doing it exclusively on OpenAI’s premium stack encourages laziness. Real winners build routers from day one, test across providers, and focus on architecture, not raw spend volume. Post-PMF distillation? Cute, but the post even admits your moat can’t be a fine-tune because the next frontier model resets everything. So you’re giving away equity for temporary acceleration that smart teams achieve cheaper through discipline. 4/ Incentives are completely misaligned. OpenAI wants you tokenmaxxing their models to justify their valuation and gather data. YC gets deal flow and hype. Founders get distracted chasing metrics that don’t equal revenue. We’ve seen this movie: over-reliance on one provider leads to nasty surprises (rate limits, price hikes, policy changes). Many teams already report 10x+ savings by multi-vendor setups. Taking this deal signals you didn’t do basic market diligence on the AI compute commoditization happening in real time. 5/ Edge cases make it even worse. Privacy-sensitive apps, regulated industries, or latency-critical products can’t always live in OpenAI’s walled garden. If you fail to hit PMF (most don’t), you wasted dilution for nothing. If you succeed big? OpenAI owns a slice and has visibility into your secret sauce. This isn’t partnership — it’s a clever equity grab exploiting FOMO in an AI gold rush. Cash is flexible. Multi-provider credits are negotiable. Building efficient systems early creates actual moats. Founders: Run the numbers. Model your projected burn across providers. Talk to lawyers about the SAFE. Compare to just raising more cash and shopping around. This deal looks generous until you realize you’re trading ownership for depreciating tokens in a market racing to zero on price. It’s not criminal, but it’s close to the line on exploiting hype. Extremely online “tokenmaxxing” culture is peak startup theater. Focus on users, product, and unit economics instead of maximizing any single vendor’s usage. Smart teams will politely decline or negotiate hard. The dumb money is taking it as-is. What am I missing? Founders in the batch — thoughts?
@sdianahu Routing easier queries to smaller models is the one most people skip and it's literally free money on your API bill lol.
@sdianahu @garrytan Icl gimme Claude tokens and I’ll take it in an instant
@sdianahu @garrytan Thanks for extending! 2M will help so many founders!
the scarce thing is not tokens. it is knowing which mistakes are worth buying with expensive inference. cheap routing too early hides the failures you needed to see. pick the routing policy owner before you pick the router, because cost, quality, and latency become product choices fast.
@sdianahu @garrytan @sdianahu if we already applied but update our application significantly in light of this new 2M$ offer, is our application update going to be reviewed.
