How Kieran Klaassen Built a Webcam Heart Rate App with Gemini 4 Argon
Built by Kieran Klaassen · · Source: X
Result: A working webcam heart rate app built in 35 minutes on low effort
What they built
Kieran Klaassen was one of six people at Every who spent a week testing Gemini 4 Argon. His standout test was a new app that reads his heart rate off his forehead using his MacBook webcam. According to Kieran, Gemini 4 built it in 35 minutes at its low effort setting.
How it works
The post doesn't explain how the app works or what prompts and setup he used. It does share what Kieran noticed about the model while coding with it:
- Good at new builds: he calls it a solid coder, and says the low effort setting is very good and fast.
- Design: it's his second choice for design after Opus 5.5. He finds it too busy on its own, but says it follows a design closely when you give it one.
- Weak at checking itself: in 17 of his 38 benchmark runs, according to Dan Shipper's writeup, the model opened and closed its code review step 30 milliseconds apart instead of actually reviewing its changes.
- Harness: Kieran says Google's tooling feels six months to a year behind, with no cloud, no SSH and no way to work across multiple computers. Dan suggests using it through Cursor or the API instead.
The result
According to Dan Shipper, the heart rate app matched Claude Opus 5.5's quality on that test and beat OpenAI's models. Kieran's overall take is that Gemini 4 isn't better than the other top models, but it now belongs with them.
How to build something similar
- Give new models a small, concrete project you can test in one sitting, like a webcam app, so you can compare them fairly.
- Run the same task on a few models and note time, quality and how much steering each one needed.
- Don't trust a model's own review step. Check its changes yourself or have a second model review them.
- If a model is strong but its app is weak, try it through another editor or the API.
The original post
Google caught up. Gemini 4 Argon isn't better than the others, but it's with the pack. Things I noticed: - It's basically a solid coder. - Low effort is very good. It built an app that reads my heart rate off my forehead through a MacBook webcam in 35 minutes. - Effort levels are well tuned and it's fast - For design it's my second place behind Opus 5.5. Too busy, but steerable. - Give it a design and it follows it. - It doesn't check its own work as well as others - The harness feels six months to a year behind. No cloud, no SSH, no multiple computers. [Quoted post from @danshipper]: BREAKING: Gemini 4 is OUT and it's SURPRISINGLY GOOD?? I admit I was ready to trash this model qhen I saw the benchmarks but six of us at @every spent the past week testing it across coding, writing, design, and knowledge work, and were pleasantly surprsied. It’s a breakthrough if your company lives in Google’s ecosystem, and a strong option for frontier hoppers as a secondary model for certain tasks. I Here’s our vibe check: My new favorite writing partner. DeepMind hired professional writers with red pens to stamp out AI slop, and it worked. I use it to untangle structure and sharpen metaphors; it also taught me something new tracing in a few of my favorite books. It's got a natural feel for structure and detailed explanations that Astra and Opus lack. Shaky at refactors, fast on greenfield apps. At low effort on a from-scratch build, Gemini 4 built @kieranklaassen a working webcam heart-rate monitor in 35 minutes—matching Claude Opus 5.5’s quality and beating OpenAI’s models on that test. Where it cuts corners is self-verification and existing codebases. In 17 of Kieran’s 38 benchmark runs, it opened and closed its code-review step just 30 milliseconds apart instead of inspecting its changes. And on my Senior Engineer benchmark—refactoring a vibe coded slop codebase—it worked for four hours and 47 minutes and scored just 21% (versus 81% to 85% for GPT-6 Astra, Fable 5.1, Sonnet 5.5, and Opus 5.5). Not great. A 2026 model trapped in a 2025 harness. Antigravity, Google’s desktop agent app, feels six months to a year behind Claude Code and Codex—lacking cloud environments, SSH, a built-in browser, and smart permission checks (leaving you clicking constant pop-ups or using unrestricted "Turbo" mode). Net Result: I keep ChatGPT and Claude open day to day, and switch to Gemini 4 when I need an editorial partner to structure a tough essay, sharpen a metaphor, or close-read a book. Reach for it for live writing collaboration, fast low-effort prototypes, or steerable UI and dashboard design—ideally via Cursor or the API to skip Antigravity. When a model is good enough to force you to use a bad harness, it tells you something. State of Play: Google hadn’t shipped a major flagship upgrade since Gemini 3 Pro last November. For teams whose companies live in Google’s ecosystem, the model you can use inside your approved work tools finally belongs with the frontier pack. DeepMind closed the model gap; now Google needs to bring the app around it into 2026. Read our full vibe check: https://every.to/p/vc-gemini-4-argon-puts-google-back-on-the-frontier?utm_source=danx&utm_campaign=vibecheck-gemini-argon&utm_content=dan-261013
Want to build something like this?
Join the Shipping Skool community free. Paid members get live bootcamp calls every week.
Join Shipping Skool — FreeMore use cases
- How Andrej Karpathy Built HackerTube with Claude Code and Opus 5.5Developer
- How Miguel Ríos Berríos Built Recall to Search Past Agent SessionsDeveloper
- How Adonias Built a CRM in Claude Code with Opus 5.5 and 6 SubagentsDeveloper
- How Federico Viticci Used Opus 5.5 to Trim His Open Minis PromptsDeveloper