Evan Liao✓

Making videos with AI shouldn’t feel like work. No editing skills. No prompt gymnastics. Just describe the idea like you’d text a friend. Pexo understands it, suggests a few creative directions, pulls from the best video models (Seedance, Kling and others), and returns a polished video ready to post. And the fun part: Pexo now runs on OpenClaw 🦞 So you can generate videos directly inside Telegram, Discrod or WhatsApp. No new apps. No tab switching. DM me "PEXO" and I'll send you an invite.

Met up with Gavin in Silicon Valley just a few days ago, and today he joined us in person at our Beijing office to speak with the entire Pexo team. He shared great on-the-ground insights on the US market—reinforcing our conviction and getting the whole team even more fired up to go all-in on the US market. Global from day one. Let's build!

Overheard a conversation at our office today between a new grad and a teammate who's been working for about 5 years. The new grad asked a dead-serious question: "Wait... did you guys actually write documents by hand from scratch back then?" That hit me hard. In just the last 2–3 years—and especially the past 12 months—how humans work has fundamentally changed. For the new generation entering the workforce, AI-native isn't a trend; it's the default baseline. We're right at the inflection point—the window where the next wave of super-apps is being formed. The way I see it, the Agent landscape is settling into three distinct layers: Personal Agents at the top — handling everyday life and lightweight tasks for everyone. Workplace Agents in the middle — connecting docs, meetings, and general office workflows. Vertical Deep Agents at the foundation — tackling complex, high-craft domain work end-to-end, like Coding Agents (or Content/Video Agents). When a new grad can't even imagine writing a doc without AI, you know there's no going back.

2022–2026 was the golden era for AI researchers. Everyone was competing to hire the youngest, smartest, and most ambitious minds to train stronger models, push the intelligence ceiling, and chase AGI. Today, model intelligence has already met—and exceeded—what most everyday tasks actually need. My take: product managers are about to take center stage again. Understanding real workflows, designing the right interactions, and figuring out monetization will matter more than raw model gains. We're at a tipping point. User habits are taking shape, and user context is starting to compound.

Gotta keep my agents busy every moment so they grow,learn, and don’t waste my credits. Haha, keep going my boys!

Pexo templates don't just copy. They pick up the essence of the video you're referencing, then make something that matches your ask.

RT @Pexoai_offical: “I want a video… but what kind?” “I can’t explain the style I want.” Sound familiar? With Pexo Templates, just pick one that fits, fill in a few details, and get a video that’s 100% yours. #AIVideo #VideoTemplates #Pexo

Hot take: if a task is easy to eval, it's probably not worth building. Thinking about sharing how we eval complex video tasks. Anyone interested? Drop a comment 👇

Excited to announce we’ve raised over $10M in seed funding to build Pexo @Pexoai_offical . I’ve always believed every person and every organization has something they want to express. A story to tell. An idea to share. A product to bring to the world. But turning what you have in mind into something others can see isn’t easy, especially through video and images. That’s why I’m building Pexo: to give full expression to every intent. Thanks to everyone who’s believed in us so far. We’re just getting started.

@Pexoai_offical We're hiring 🎬 Looking for people who want to help everyone express themselves through video: Engineer / AI Researcher / Designer… DM me or reach out at hr@apexcontext.com

just met one of our early paying users today. a few things stuck with me: she loves Mark to Fix. no long prompts, just point at what's off. she calls @Pexoai_offical her "secret tool" and almost didn't want to share it. hahah she's super AI native and doesn't even know it. that's who we're building for.

2028: flying cars finally ship 2029: this video

feels like I don't have enough work to keep them all busy

Just had a deep convo with Gavin. As AI founders, we focus way too much on SF, NYC, LA. Houston, Phoenix, Atlanta are growing fast. Housing, elder care, healthcare are full of unmet needs. Go deep. Go vertical. Honestly a wake-up call.

Your personal video agent.

Now you can create videos the way you vibe-code. Meet Pexo, your AI video agent. Share your vision and references. Work through the details with Pexo like you’re messaging a friend. No complex prompts. No new tools to learn. Want to change something? Mark it on the video and tell Pexo what to change, just like leaving a comment in a doc. Everything in this video was made with Pexo: the visuals, the motion graphics, the music, the captions, and the voiceover. Even our founder @evanLiaoQ appears on screen. You vibe-coded what you built. Now vibe-create how the world sees it. #Pexo #AIVideoAgent #VibeCreate #MadewithPexo

Conversations with founders from Monid @shengkunye and Photon @danieltian , plus a visit to Exa, left me energized. An idea can start with a joke, some fun, or a problem of your own. Exa reminded me what comes next: staying committed to a vision, even when the path gets hard.

Your brief can be a vibe. Even "black, but full of color." Just say it. Mark it. Pexo brings it to life.

Yahoo solved an early-internet problem: people didn't know how to find websites, so it gave them a directory. But a UI can only show so much, so recommendation algorithms took over and got you to the right site faster. Today's AI aggregators have obvious value too. Instead of subscribing to N different products, you just use one aggregator. So what's next? Does the aggregator start solving your problem for you, or does a general-purpose agent make aggregation unnecessary?

me doing my 4th launch dance this quarter pexo: still filming like it's the first one

you cannot out-shitpost me (with the help of muse wonder full)

After a session on how Clay grew, I gave its website to Pexo and asked it to make a video. Here’s the result. We’re building Claude Code for video creation. This is one example of what you can do with it.

Went to a Voice Coding hackathon during Silicon Valley Tech Week today. My biggest takeaway wasn’t about coding. It was about when and where we can use AI. I’m literally having this conversation with AI by voice while sitting in a car. Ideas, requests, status checks. Things that used to require sitting down at a computer can now happen anywhere. Voice isn’t just a faster way to input information. It removes the need to be at your desk in the first place. That could change how often we use AI — and what we use it for.