Radish: The FriendLLC Model

Phil Levin's published two-entity legal and governance structure for Radish, a 20-adult, 8-child co-living compound on a third-acre lot in Oakland, California. A landlord LLC owns the property, the resident community governs daily life, and the same person can be both an owner and a resident with different decision rights in each role.

Radish is a 20-adult, 8-child co-living compound on a third-acre lot in the Temescal/Longfellow neighborhood of Oakland, California. It was founded in 2019 by Phil Levin, Kristen Berman, and six other initial residents. The site holds 6 buildings with 10 housing units. As of 2024 there are 19 owners and 13 residents, with substantial overlap between the two groups.

The constitutional design is the FriendLLC model. There are two entities and two governance layers:

The first entity is Big Cabbage LLC, a Real Estate Limited Partnership that owns the property. Owners (people who put in capital) hold shares in this LLC and make property-level decisions: when to sell, when to refinance, how to spend the maintenance budget. The LLC pays property tax, holds insurance, and handles all standard landlord functions.

The second entity is the Radish Community, the current residents. They make daily-life decisions: how the food system works, who gets to live there, how rent is split between residents.

A single person can wear both hats. Phil is both an owner and a resident. Arthur is an owner who does not live there. Suzannah is a resident who has not invested. The structure handles all three cases.

Money flows: Each resident pays rent based on the size and quality of their unit. The community uses some of that for shared expenses (food, internet). A property rent is paid up to Big Cabbage LLC, which covers property tax, insurance, maintenance, and pays a dividend back to owners. Radish targets a 5% annual return on dividend (a normal middle-of-the-road real estate cap rate), which keeps the project legible to investors and lenders without overpaying.

Owner governance: A Manager handles small day-to-day decisions (Phil holds this role at Radish, estimated at 1 to 2 hours per month). Larger decisions go to a vote, with thresholds escalating by stakes. The published Duck Cloud voting threshold table (a separate Phil-led project) is the worked-out reference for how to set those thresholds.

Resident governance: Decided by the residents according to whatever process they adopt. At Radish this includes regular community meetings and explicit rules for who joins.

The strengths: Flexibility (the model adapts to changing owners, residents, and property without rewriting the agreement), accommodation of unequal financial means (some friends invest a lot, some none, the structure handles it), legibility to outside parties (banks, accountants, insurers all understand a Real Estate Limited Partnership), and limited liability for owners.

The weaknesses: Exit is slower than direct individual ownership (exits happen through resale of the property, refinancing, or new investors buying out old). Tax inefficiency (resident-owners pay rent to themselves and get taxed on it). Administration overhead (someone has to do the books).

For a researcher, Radish is the case where a co-living community published its full constitutional design as a how-to document, with the explicit intention that other groups would copy and adapt it.

Living examples and entry points for community builders

If your community is trying to figure out how to share ownership of a home without forming a full co-op, the FriendLLC model is one of the more practical experiments happening right now.

Radish (Brooklyn, New York) is a coliving house co-owned by roughly eight people using an LLC operating agreement written specifically for shared residential ownership between friends. The structure is designed to be legally simple enough that a small group can set it up without years of nonprofit formation, while still being robust enough to handle the hard questions: what happens when someone wants to leave, how decisions get made, who can block a sale. It is a live experiment, not a finished model, and the people running it are open about that.

Dreamship (documented separately in this library) developed a parallel LLC structure for shared ownership of a sailing vessel, which later influenced how some small property groups think about co-ownership. The overlap between the two models is significant enough that people working on FriendLLC-style arrangements often end up reading Dreamship's operating agreement alongside Radish's documentation.

Across the US, a small number of coliving LLCs have replicated elements of this pattern without coordinating with each other. They are not a network yet, but the convergence suggests that the FriendLLC structure is filling a real gap between informal household sharing (no legal structure) and full housing co-ops (significant governance overhead).

For adjacent models, community land trusts accomplish some of the same goals (keeping housing affordable over time, preventing speculative resale) but require more organizational infrastructure and usually involve a nonprofit entity holding the land. Housing co-ops have democratic governance built in but come with more bureaucratic overhead than most small groups of friends want to take on in year one. The FriendLLC sits between these: more formal than a handshake, lighter than a co-op.

Where the door is:

• Radish has posted documentation of their model online. It is worth reading before you talk to a lawyer, so you know what questions to ask.
• The most important conversation to have before structuring anything: what does your group do when one person wants to sell and the others don't? Work through that scenario in plain language before you open a document.
• Find a lawyer who has worked with intentional communities or shared ownership before. General business attorneys often default to structures that don't account for the social dynamics of co-ownership between friends.
• Ask your group directly: how much legal formality matches the level of trust we actually have right now?

One caution: LLC law varies significantly by state, and an operating agreement that works well in New York may create unexpected problems in California or Texas. Do not copy a template directly. Use it as a starting point for a conversation with local legal counsel.