Macrite Is Now Part of the Google for Startups Cloud Program

A small update I'm happy to share: Macrite is now part of the Google for Startups Cloud Program, Google's program supporting early-stage startups building with Google Cloud.

Less than a year ago, Macrite was an idea on a whiteboard: what if a nutrition app could actually think? Not just count what you ate, but help you decide what to eat next. Today it's a live app, available in three languages.

Why It Matters

Macrite is bootstrapped and built by a solo founder. Every part of it, from the scan that has to come back in seconds to the servers that keep it running day and night, is something I build and pay for myself.

The program is designed for bootstrapped companies like Macrite, and it helps with exactly that. The credits and resources let me keep Macrite fast and reliable as more people use it, without cutting corners, while still putting more into the features you use every day.

What It Unlocks

Being part of the program gives Macrite:

• Google Cloud credits.

  More of Macrite's budget goes straight into the product.
• Architecture workshops.

  Sessions with Google's startup team on building and scaling the right way.
• AI and machine learning resources.

  Access to Gemini and the Vertex AI platform, to explore what's next for Macrite's AI features.
• Training.

  Technical onboarding and courses through Google Cloud Skills Boost.
• Startup events.

  A chance to learn from other founders building on Google Cloud.

Where This Meets Macrite

Macrite doesn't lean on one AI model, or one company. Each job goes to whatever handles it best, and a surprising amount happens right on your phone.

• On your phone.

  Your calorie and macro targets are calculated on the device from published formulas like Mifflin-St Jeor, not guessed by an AI. Voice entry uses your phone's built-in speech recognition. And on supported devices, AI models running on the phone itself spot the foods in your photo when you review a scan.
• Vision in the cloud.

  Cloud vision models identify what's on your plate, on a restaurant menu or in your fridge, and estimate portions.
• Reasoning in the cloud.

  Reasoning models turn that into advice: the best dish on a menu for your remaining macros, recipe ideas from your pantry, and Mac's Corner suggestions for your next meal.
• Macrite's own logic.

  Before any AI writes a suggestion, Macrite's own rules work out what matters right now, like how much protein you have left or which meal is next.
• Real data, not just AI.

  Barcodes are looked up in public food databases, and pantry recipes come from Macrite's own curated recipe library, ranked by how well each one fits your pantry and your macros.

Google's AI resources fit into this nicely. They give me room to test newer vision and reasoning models as they arrive.

But Macrite was built so that no single model or provider is irreplaceable. Each task can use whichever model does it best, on the phone or in the cloud, and switch when something better comes along. That keeps the focus where it belongs: scans that get more accurate, and suggestions that get more useful.

What It Means for You

Nothing changes in how you use Macrite. Snap, speak or type a meal and get your calories and macros in seconds, same as before. I just have a little more room to keep improving it.

And your personal details stay yours: they're never sold or shared with advertisers, and your name, email and account are never shared with AI providers.

What's Next

I just launched Macrite in German and Spanish. Android is on the way. More languages are coming. And I'm building smarter ways to plan your meals around the day you're actually having, not the one a spreadsheet expects.

Thank You

Thanks to everyone who tried Macrite early, sent feedback, or told a friend. A lot of what's in the app today came from those messages.

If you haven't tried Macrite yet, now's a good time to give it a go.

Ideas for what I should build next? Tell me at support@macrite.app. I read every message myself.