▶ LIVE
VRChat hits 100K+ concurrent on Steam, again Series D total: $80M, led by Anthos Capital $96M raised across a decade Founded by two Oculus DK1 obsessives Worlds, avatars, Udon - all built in Unity San Francisco HQ, remote-first team of ~130 VRChat hits 100K+ concurrent on Steam, again Series D total: $80M, led by Anthos Capital $96M raised across a decade Founded by two Oculus DK1 obsessives Worlds, avatars, Udon - all built in Unity San Francisco HQ, remote-first team of ~130
YesPress Profile / Company / Gaming & Social VR

VRChat Inc.

The social network you have to put on your face. A platform of millions of worlds, almost none of which it actually built.

Founded2014 HQSan Francisco Funding$96M+ Team~130
VRChat logo
A logo built for headsets, captured here in two dimensions under duress.

Dispatch · San FranciscoIt is 9:14 p.m. somewhere, and 100,000 strangers are hanging out.

A blue-haired anime character is teaching English to a chess-piece avatar in a Tokyo train car that does not exist. Two doors down, in a world someone built last weekend, a public speaking club is rehearsing toasts. A medical resident is using a virtual cadaver lab. Somewhere else, a guy who looks like a sentient burrito is DJing. This is not a metaphor. This is Tuesday on VRChat.

VRChat is a free social platform that runs on PC, on Meta Quest, and on whatever middleware Unity ships next. It is one of the most populated places in virtual reality. It is also one of the strangest companies in San Francisco - a 130-person operation whose product is mostly built by people it does not employ.

VRChat is the rare metaverse that didn't need a Super Bowl ad to fill up.
- YesPress observation, May 2026

The ProblemThe Internet Forgot How to Be a Place.

By the early 2010s, the social internet had finished its great flattening. Feeds replaced rooms. Photos replaced presence. You could broadcast yourself to thousands of strangers, but you could not really sit next to one of them. The web had become a great noticeboard with very poor furniture.

Then in 2012, a Kickstarter campaign for a clunky head-mounted display promised to put you inside a screen. Most people who funded the original Oculus Rift wanted to play games. A few of them wanted something stranger: a place to go. A place that wasn't a feed.

The web is a library. VR is a city. You can read a library. You can only live in a city.
- Paraphrased from VRChat's own pitch decks

The BetTwo strangers, one Reddit thread, one DK1.

Graham Gaylor backed the Oculus Kickstarter early and got one of the first DK1 prototypes shipped in 2013. He wanted to use it for something nobody had built yet: a place to meet people. He wrote a rough prototype called VRChatroom and posted it to Reddit looking for testers.

Jesse Joudrey had founded a small VR Unity outfit called Jespinage in 2013. Graham heard Jesse on a podcast describing the exact features VRChat needed. They talked. They agreed. They incorporated VRChat Inc. in January 2014 and released the first build for the Rift DK1 on January 16th of that year. It was lo-fi, janky, and entirely earnest.

Graham Gaylor
Co-Founder · CEO

Backed Oculus on Kickstarter. Wrote the first VRChat prototype. Now spends his days arguing about creator monetization.

Jesse Joudrey
Co-Founder · CTO

VR Unity engineer. The technical conscience behind avatars, worlds, and the company's homegrown scripting language.

They didn't pitch a metaverse. They built a chatroom, then let it eat the metaverse's lunch.
- YesPress profile, May 2026

Twelve years, abridgedFrom DK1 to 100K concurrent.

2014
VRChat Inc. founded. First DK1 release.
2017
Steam Early Access. Goes viral on YouTube.
2018
Series B, Udon language, world creators take off.
2021
$80M Series D led by Anthos Capital.
2024
Concurrent records broken. Marketplace expands.

The ProductA toolkit pretending to be a hangout.

To users, VRChat is a hangout. To creators, it's a software stack. The platform ships a Unity-based SDK that lets anyone publish a custom avatar or a custom world. Inside those worlds, behavior is wired up with Udon - VRChat's in-house visual scripting language, named after Japanese noodles, because of course it is.

The toolkit is the trick. Avatars get full-body tracking, lip sync, eye tracking, hand gestures, and physics on hair and clothing. Worlds get spatial audio and persistent rooms. A creator can build something on a Saturday and ship it to millions of headsets by Sunday. Big-budget platforms have spent billions trying to clone this loop. They mostly haven't.

What you can do here Build a club, host a stand-up night, run a language exchange, attend a wedding, rehearse a TED talk, attend an avatar fashion show, fall asleep in a virtual cabin while someone reads to you. People do all of these. Last week.
What the company actually builds The runtime, the SDK, the moderation backend, the trust system, the cross-platform networking layer, and Udon. Then they get out of the way and let the worlds tab do the rest.
Most apps want you to consume content. VRChat wants you to host the party.
- YesPress field note

The ProofThe numbers, briefly, with citations on request.

Critics of social VR have argued for a decade that nobody actually shows up. Then somebody opens Steam Charts. VRChat's all-time concurrent peak crossed 130,000 in 2024 - on Steam alone. That excludes the Quest standalone users, who are roughly half the audience.

VRChat by the numbers

Sources: public Steam data, company disclosures, press
Peak CCU130K+
Series D$80M
Total raised$96M+
Revenue est.~$28M
Team~130
130K+
Peak Steam concurrent
$96M
Total funding
10+
Years operating
User-made worlds
VRChat doesn't have engagement. It has tenants.
- YesPress, on the difference between scrolling and showing up

The MissionAnyone, any device, any world.

VRChat's stated mission is to let anyone create, share, and play in virtual worlds together, regardless of device. That last part matters more than it sounds. A Quest user, a PC VR user, and a desktop user with no headset at all can occupy the same room and see one another move. Cross-platform parity is not a marketing line in this company. It is an engineering religion.

The business model is still settling. VRChat Plus subscriptions cover some of it. A creator economy with an in-app currency and marketplace is being built out for the rest. CEO Graham Gaylor has been on the podcast circuit lately making the case for UGC monetization that pays creators meaningfully - not the platform-takes-most arrangement most app stores prefer.

The question isn't whether VRChat is a metaverse. It's whether the metaverse will keep being VRChat.
- YesPress

What's NextWhy this matters tomorrow.

The big platform money in VR has been chasing the wrong question. Bigger headsets. Better avatars. Sharper hands. VRChat has been quietly answering a different one for twelve years: what do people actually do when you put them in a room together and don't give them anything to do?

It turns out they teach English to chess pieces. They host stand-up nights. They build virtual cadaver labs. They rehearse toasts. The platform's wager - that user-generated worlds beat studio-built experiences - is starting to look less like a niche bet and more like the only bet that works.

And so we return to 9:14 p.m. somewhere. The blue-haired anime character has finished her lesson. The chess piece thanks her in halting English. They both log off, eventually. The virtual train car keeps running, empty, waiting for the next two strangers. VRChat doesn't need to fill it. Somebody else will.

The metaverse is not coming. It's been here since 2014. It's just wearing a furry costume.
- YesPress closing argument

Watch & listen

Find VRChat

Share this profile

LinkedIn Twitter / X Facebook Instagram