Bumble BFF, Timeleft, Meetup, We3, Boo — friendship apps in 2026

I tried roughly every major friendship app in the U.S. over the past 18 months as research before building Are We Friends?. What follows is the honest comparison — what each one does well, where each one fails, and what's still missing from the category.

Disclosure upfront: I built one of the apps in this list. I'll flag where my view is biased and try to compare like-for-like.

Bumble BFF

The category leader by user count. Built on the same swipe-deck UX as the dating product, with a "friendship mode" flag flipped. Available in essentially every country Bumble operates in.

What it does well: - Pure scale. If you're in a major city, Bumble BFF probably has more candidates within your radius than any other friendship app. The matching pool advantage is real. - The brand recognition. Saying "I matched with someone on Bumble BFF" is socially normal in a way "I matched on Patook" isn't. Lower stigma than smaller apps. - The UX is polished. It's been A/B tested on millions of users. Every interaction feels smooth.

What it does badly: - It's a dating UX with the dating goal removed. Swipe, match, 24-hour message timer, "hey," dead thread. The funnel leak we wrote about here is most visible on Bumble BFF. - Matching is shallow. Photos and a short bio. No personality test, no facet matching, no real signal beyond surface compatibility. - No first-meeting plan. You match, you say "hey," you small-talk for two weeks, you never meet up.

Verdict: best for the matching pool, worst for actually getting to a coffee meeting. If you have lots of patience and a high tolerance for "hey what's up" loops, it works. If you don't, it doesn't.

Timeleft

The category's most interesting newcomer. The model is six strangers at a Wednesday-night dinner, matched by a short personality quiz, with a night-before reveal of who else will be there.

What it does well: - Removes the planning problem. You don't have to figure out where to meet — the platform picks the restaurant, the time, the group. You just show up. - Built-in conversation game. Timeleft includes a printed conversation-starter game at the table, which dramatically lowers the small-talk activation energy. - Group dynamic. Six-person groups are an interesting middle ground between 1-on-1 (which can feel like a date) and large meetups (which dilute connection). - Strong brand and PR. Lots of media coverage. People talk about Timeleft.

What it does badly: - One-shot, not relational. You meet six people on Wednesday. By Friday you've either swapped contact info or you haven't, and Timeleft makes no real effort to convert dinners into ongoing friendships. The ongoing-relationship layer is just... missing. - Wednesday-only, by design. Cool when it fits your schedule, useless when it doesn't. No 1-on-1 option, no other days. - Matching is shallow. A short quiz produces a category bucket; you don't actually know your match's personality profile in any specific way. - Pricey per outcome. Around $22/dinner plus a membership fee. If you go to four dinners and one becomes an ongoing friendship, that's ~$100 per friendship — high but not crazy.

Verdict: great as an ongoing social activity for someone who wants strangers-on-a-Wednesday-night dinners. Limited as a friendship pipeline because the conversion rate from dinner to ongoing friendship isn't built into the product.

Meetup.com

The OG. Founded in 2002, still alive, mostly used for hobbyist groups (running clubs, book clubs, hiking, technology meetups).

What it does well: - Activity-first. You search for a hobby, you find groups around it. The recurring-activity model is the right model for friendship formation in the abstract. - Free for attendees. Organizers pay; attendees don't. - Genuinely active in some niches. If you're into a specific technical hobby (board games, climbing, certain programming communities), Meetup is often where the real-world community is.

What it does badly: - The platform itself feels abandoned. Old design, slow updates, organizer churn means many groups list events that don't actually happen. - No matching at all. You show up to a hike. You meet whoever shows up. There's no pre-filtering for compatibility — just topic interest. - Open RSVPs lead to no-show problems. Meetup events routinely have 30% no-show rates. The social cohesion of the group is consequently weak. - Built for organizers, not attendees. Event hosting tools are decent; friendship-development tools don't exist.

Verdict: best as a finder for recurring real-world activities. Bad as a "friendship app" — the friendship part is entirely on you.

Hey! Vina (RIP)

Folded in 2022. Was the most-cited "Bumble BFF for women" alternative. Lessons from the failure: the women-only positioning was useful but the underlying UX was still swipe-deck-and-message, which has all the small-talk-gap problems above. No amount of better community-vibes copy fixed the fundamental funnel leak.

We3

Three-person matching with a personality quiz. The pitch: trios produce stronger ongoing dynamics than dyads.

What it does well: - Group-of-three is interesting. Threes have different social dynamics than dyads — less pressure on any one relationship, more sustainable as ongoing groups. - Matching has real personality content. Not as deep as a full Big Five profile but better than Bumble BFF.

What it does badly: - Three people are hard to match well simultaneously. The combinatorics make the match pool effectively smaller. In some cities you might wait weeks for a high-quality trio match. - Coordinating three schedules is harder than two. Group meetups have logistical friction that dyads don't. - No first-hang plan. Same small-talk gap as Bumble BFF, but now with three people who have to all overcome it.

Verdict: interesting model, but the trio framing might be a constraint that doesn't map onto how friendships actually grow.

Boo

Personality-matching app explicitly built around MBTI types. UI is clean, the matching is presented as personality-driven.

What it does well: - Personality-first matching is the right idea. Putting personality at the center is a real improvement over photos. - MBTI is familiar. Most users have already taken some version of an MBTI test, so onboarding feels lighter.

What it does badly: - MBTI itself is academically discredited. Test-retest reliability is bad (50% of people get a different type when they re-take), the binary categorical structure isn't supported by the data (continuous Big Five-style dimensions fit better), and the four dimensions aren't truly independent. - The matching model is basically "find people with similar 4-letter codes." Shallow at the matching layer in ways that compound when the test underneath is shaky. - No first-hang plan, encryption, or post-match infrastructure. Same gaps as everyone else.

Verdict: good first step in the right direction (personality matching) on top of a flawed instrument (MBTI). Big Five is meaningfully better than MBTI for actual matching, and Boo's UI would be better if it switched.

Patook (RIP)

Strict-friendship-only app that aggressively banned anyone who treated it like a dating app. Folded around 2023. Lessons from the failure: enforcing the platonic-only norm was useful, but the matching was still shallow and the post-match flow was still "hey." Same funnel leak.

NextBestie

Newer app. Personality quiz + interest tags + city-level matching. UI feels modern.

What it does well: - Personality quiz is lighter than Big Five but more substantive than Bumble BFF. - Clean UX, mobile-first.

What it does badly: - Small user base. Match pool in many cities is thin. - No first-hang plan. - No encryption story.

Verdict: probably worth trying if you're in NYC, LA, or a similarly dense market. Limited in smaller cities.

What's missing from the category

Looking across all of these, there are three things consistently missing:

1. A real personality test, used consistently. Big Five with 30 facets, not MBTI 4-letter codes, not 12-question quizzes. The depth of matching matters.

1. Plan-the-hang built into matching. None of the apps above pre-draft the first-meeting plan. They all leave the user to write "hey" and hope.

1. End-to-end encrypted DMs. Friendship conversations get personal. None of the major apps do real E2EE. Signal-style cryptography for non-romantic messaging is genuinely uncommon.

I built Are We Friends? around exactly these three missing pieces. Bias disclosed: my judgment of the gap is informed by what I built. But the gap is real, and you'd have noticed it independently if you'd tried all the apps above.

Recommendations

If you're picking a friendship app today:

• Best for "I want a big match pool": Bumble BFF. Despite its UX problems, scale is a real advantage.
• Best for "I want a planned in-person experience": Timeleft. Stop expecting it to produce ongoing friendships, treat it as a recurring social activity.
• Best for "I want a recurring activity": Meetup. Pick the hobby, show up consistently, accept that the friendship part is entirely on you.
• Best for "I want depth-of-match + first-hang plan": Are We Friends?. Yes, biased.
• Worst pick: any pure swipe-deck friendship app without personality matching. The funnel leak will eat you alive.

The category is still young. Two years from now this list will look very different.