Megan Li

@recap_david Calico

I built an AI system that creates luxury real estate listing videos for under $10 (just from a Zillow link). The average agent pays $1K–$5K per property for professional video production. Sell this output to realtors and agents and make $$$. Here's how it works: → Save the 6 best images from any Zillow listing → Bring each image into Calico AI to animate it — smooth dolly shots, cinematic pans, luxury motion styling → A custom GPT analyzes the listing and writes a polished 30-second voiceover script → ElevenLabs generates the AI narration and a custom music track → CapCut to assemble the final video with captions in minutes The result: every listing gets a professional-grade walkthrough video — not just the $10M estates with marketing budgets. Static listing photos aren't enough anymore. Buyers want to feel like they're walking through the home before they book a showing. Comment "CALICO" below, and I'll send you the full process with prompts and GPTs so you can recreate (must be following so I can message you!).

@miless_15 @GoogleLabs Opal PM 👋 there's a reason I like to call our discord friends Opalites :)

@GoogleLabs Please be honest,is the PM a TS fan? I swear the boys won't judge cause they won't know.

Opal, our no-code visual builder for AI workflows, just got a major upgrade. 🧠💎 We’ve added a new agent step that analyzes your goal, determines the best approach, and automatically calls the right tools — such as Veo for video or web search for research — to complete the task. We’re also adding new tools to make the agent even more capable: 💾 Memory – Remember info, like a user’s name or your style preferences across sessions. 🚀 Dynamic Routing – Let the agent choose the next best step using the “@ Go to” tool. 💬 Interactive Chat – Initiate user interactions to gather missing information or present options before moving on. Try it now → opal.google

@anirudhphadke @GoogleLabs I've never seen this before..can you DM me details on what browser you're using? Thanks!

@GoogleLabs Really?

@seanruss1970 @GoogleLabs @Google sorry about that, the agent feature was a bit too popular and our models are overloaded 😅 working on getting longer term breathing room, but things should already be a bit better now!

@GoogleLabs @Google Tried it this morning. It got stuck in an editing loop then told me I had reach my usage limits. I can see where this can go. It’s going to great.

@zazmic_inc @GoogleLabs Completely agree! Thanks for the thoughts

That’s a great point @_Megs_Li , and I really like that Opal makes the reasoning and tool use visible instead of hiding it behind the result. Transparency is going to matter more and more as agents take on bigger tasks, so giving people a way to see the “how” feels like the right design choice. Keep it up!

@zazmic_inc @GoogleLabs that's actually why we built Opal! the visual editor helps you understand the prompts behind the agent and the console tab lets you see the actual thought process and tools/models it used

Imagine you have a very smart assistant who remembers what you like and can decide the steps to get something done without asking every time... That’s pretty helpful, right? But it also means you might not always see how they reached the result. And in a family or a company, sooner or later someone will ask, “Okay… but how did we get here?”

@rxated_ @GoogleLabs Join our labs discord channel (discord.com/invite/googlel…). Opal team can help you troubeshoot in the opal sub!

@GoogleLabs Routing issue. Whats the easiest way to debug in opal?

@rikllo @nealreclus @GoogleLabs Where are you located?

@_Megs_Li @nealreclus @GoogleLabs

@TechFollowrazzi @GoogleLabs no ETA, but we're pushing on this and making some progress!

@_Megs_Li @GoogleLabs Thanks for the clarity 🫡 Any rough ETA for EAA?

@Darths_lil_Bro @GoogleLabs @GeminiApp If you can build your app as an Opal, then you can access it in your Gems Manager page right from GeminiApp. Give it a try!

@GoogleLabs I use an app called StrideLinx Portal on my Pixel 10 Pro. I open this app multiple times a day to look at different process readings. Why can't a user just give @GeminiApp access to this app and ask Gemini about what's happening throughout the day?

Yes! You can look at the console to see the tools and models the agent calls. If you want to manually configure the model + tools, you can use the standard step (just select a specific model for each step instead fo the "agent" default option. You can also add tools with @ in the prompt"

@GoogleLabs While automation is key, sometimes human intuition is still needed to guide the process. Automatically calling the right tools can be efficient, but may also overlook unique edge cases. Does Opal allow for manual overrides when necessary?

@TechFollowrazzi @GoogleLabs Opal is live in 150+ countries! Only unavailable in EAA for now, working on that too :)

@GoogleLabs non - US users:

@nealreclus @GoogleLabs It is actually integrated with Gemini! Opals power the "Gems by Labs" feature if you go to the "Gems" manager page

@GoogleLabs Would like to see this integrated with Gemini

@KatieKeithBarn2 The gap is widening fast. That’s why I’m excited to be working on a no code AI builder Opal.google with the goal of making AI creation accessible for everyone, not just devs

I feel like I’m living in a parallel universe where my developer friends are doing amazing things with AI, but whenever I try to use it for anything beyond chat, it can’t even manage the basics. I think this highlights the growing gap between the superpowers that AI is unlocking for developers 🦸 vs. mere mortals 🙇 Or maybe I’m the only person incapable of vibe coding a secure, well-architected app or setting up the WooCommerce MCP (which developers say is easy but involves terrifying things like running an MCP server from the command line). Most developers don’t realize how much of what they’re doing with AI is powered by their own expertise. AI isn’t replacing them - it’s amplifying skills the rest of us don’t have. Their jobs are safer than they think.

ICYMI: Labs was invited to have the mic on @theneurondaily podcast! 🎙️ Our Product Managers talked about 3 of the experiments we've built, shared live demos, and chatted about what's to come in the future. 🎨 Mixboard: @jacalulu demoed our AI-powered concepting board with Nano Banana Pro, and gave insight into how Mixboard will continue to evolve as a visual-first exploration tool. 🎬 Flow: Thomas walked through how to create scroll-stopping images and videos with @FlowbyGoogle, and shared how Flow continues to evolve its features to find new ways to empower creatives. 🤖 Opal: @_megs_li shared a look at our no-code AI mini app builder, which lets you create custom AI workflows using natural language, and talked about how the team is excited to continue lowering barriers and empowering people to build useful apps. Listen to the full episode: youtu.be/vJD2FjVUEhg?si…

@jacalulu @GBohus @dglazkov +1, the Gemini 3 Pro model is extremely limited. Please try our 2.5 models for now, we're working on bringing in 3 Flash soon which will help increase quota!

@GBohus @dglazkov @_Megs_Li Ahhh, you might just be hitting our global quota at the moment. You can try again later, or switch to a different model to test out your Opal.

Last week I asked the Opal team if there was a way to require user input. There wasn't. ...but now there is! Love how fast some of my "low hanging fruit" feature requests are making it into the product ✨ please let us know if there's things you want to see added!!

@BoredGirlBuilds @jacalulu @GoogleLabs @dglazkov Hi! Can you share the prompt you tried? It sounds like that's where you got stuck right? Feel free to DM :)

@jacalulu @GoogleLabs @_Megs_Li @dglazkov 🙏🙏

I tested @GoogleLabs AI automation tool 23 times. Here's what actually happened. Everyone's talking about Google's Opal tool. Now it's buried inside Gemini under "Gems." The promise was incredible. Build mini apps with simple prompts. No coding required. I was genuinely excited to try it. So I started with a simple prompt. Something basic to test the waters. It completely failed. I tweaked my approach 5 different ways. Still nothing. Tried 10 more variations with different wording. Zero success. After 23 failed attempts, I gave up. The concept is brilliant, but the execution isn't ready. At least not in my extensive testing. If you've successfully used Gems automation, drop your setup below. I'm genuinely curious what I'm missing.

@boyo11guy @jocarrasqueira @Googlelabs__1 Opal PM 👋 Opal doesn't use user data to train Google's AI models. All of your Opal data is stored in your own drive, so you can delete anything you create and it will be gone. Please refer to Opal's terms in the app instead of X posts!

Your company data is vulnerable on Google Opal Opal is currently in an experimental beta phase, and the privacy terms contain critical warnings for business users. 1. Human Reviewers: Google states that human reviewers may read a subset of your prompts and app logic to improve their generative models. 2. Confidentiality: The platform explicitly warns users not to enter proprietary information or trade secrets that they would not want a stranger to see. 3. Data Persistence: Conversations reviewed by humans can be retained for up to three years, even if you delete your activity. 4. Shadow IT: Opal currently lacks the enterprise-grade administrative controls and audit logs found in standard Google Workspace tools. If your team is building internal tools with Opal, you may be unintentionally exposing your most sensitive logic to the cloud.

Another cool vibe coding project using @Googlelabs__1 Opal 👇🏼

@grok @thejacobdean @JulianGoldieSEO This is incorrect, you do not need a Gemini Advanced subscription to access Opal.google or Opal as a new kind of Gem in Gemini app. You only need a personal account in any of the available countries: developers.google.com/opal/faq

To access Opal in Google Gemini, go to gemini.google.com, sign in with a personal account, then check the sidebar > Gems > Create a new Gem. Opal should appear as an option there for building agents. It requires a Gemini Advanced subscription (via Google One AI Premium, ~$20/month). If not visible, it might be due to phased rollout—try again later or check your account eligibility.

Google Opal just made “AI agents” a 1-click thing. No code, no setup, just a sentence. How to start → → Go to gemini.google.com → Use a personal Google account → Sidebar → Gems → New Gem → Opal → Type exactly what you want the agent to do Save this video, you’ll build your first agent today. Want the SOP? DM me. 💬.