Thoughtfully made for your Mac

Bonsai

Local AI, made to fit.

Find the right AI models for your Mac.
Know what fits before you download.

Skip introduction

Made for the Mac you have

A clear answer.
Before the download.

Know what will run well on your Mac, and why. Bonsai turns your hardware and your task into a thoughtful next step.

See how it works
Bonsai’s recommendation screen showing a fit verdict, estimated speed range, and memory headroom. Illustrative preview data.
Actual app interface · Illustrative preview data.
Your machine. Your task. Your best fit.01 — RECOMMEND

Less trial. Less error.

Your Mac can run local AI.
Finding the right model
shouldn’t be the hard part.

Model sizes. Memory limits. Quantization. You shouldn’t need an afternoon of research to get started. Bonsai turns what your Mac can do into a clear next step.

01 / A considered recommendation

One machine. One task.
One clear answer.

Built around your hardware.
Informed by what you want to do.

Step 01

Start with what you have.

Bonsai reads your chip, memory, and available storage. It builds a picture of your machine before a single model is downloaded.

Know what fits before you download.
Bonsai hardware scan showing an Apple M4 Pro with 24 GB memory. Illustrative preview data.
Actual Bonsai interface · Illustrative preview data.

02 / Install & validate

From “will it run?”
to ready to go.

Install through Ollama, then run a short check on your Mac. Compare what was predicted with what actually happened.

No second-guessing the download. No hunting for the next step.

Installed through Ollama. Validated by Bonsai.
Bonsai’s completed installation screen showing an As expected benchmark result. All measurements are preview data.
Actual Bonsai interface · Illustrative preview data.

03 / A little order

Your local AI library,
finally understood.

Different tools. Different folders. One clear picture of what’s on your Mac.

Bonsai Models interface showing storage across Ollama, LM Studio, and Hugging Face, followed by installed artifacts. Preview data.
Actual Bonsai interface · Illustrative preview data.
01

See the whole shelf.

Discover models across Ollama, LM Studio, and the Hugging Face cache. Keep their sources in view.

02

Understand the space.

See what’s stored and where. Shared files are counted once, so the picture stays honest.

03

Make room, carefully.

Verify matching copies by content. Review supported cleanup actions before anything is removed.

Your say, always. Nothing is deleted automatically.

AT HOME WITH YOUR LOCAL TOOLS
OllamaLM StudioHugging Face
Actual Bonsai menu-bar interface showing the active runtime, a Running well state, speed, memory, and context. Synthetic preview readings.
Actual menu-bar interface · Preview readings.

There when you need it.
Quiet when you don’t.

04 / Quietly keeping watch

A little reassurance.
Right in your
menu bar.

Keep an eye on your active model and the readings your runtime exposes. Bonsai puts useful context close at hand, even when the main window is closed.

Available readings vary by runtime. Missing metrics stay missing.

05 / Room to grow In development

A better fit.
When it really is one.

Bonsai is being built to revisit your saved task as its model catalog improves—and let you know when a meaningfully better fit comes along.

A useful nudge. Only when there’s a reason.

Catalog updates are not active in the current build.

The Bonsai way

Powerful things thrive in the right space.

Room to grow.
Space to think.

Local AI doesn’t have to mean running the biggest model. It means finding the right balance of capability, speed, and the machine you already own.

Close-up of weathered, moss-flecked bark where a trunk divides into branches.

Your Mac. Your models. Your business.

Bonsai’s core work happens locally. Your model files, prompts, and inventory aren’t sent to a backend for basic operation. The app has no analytics transport.

This preview site includes an early-access signup form (provided by Kit). The hosting service may process standard requests and access logs.

Pricing

Start free.
Keep your local AI in shape with Pro.

Find the right model for free. Upgrade when you want Bonsai to manage, monitor, and maintain your local AI setup.

Bonsai Free

$0

Find out what local AI actually fits your computer.

Download Free

No credit card required.

  • Hardware scan
  • Task-based model recommendation
  • Predicted speed and memory fit
  • Top model recommendation
  • Installed-model detection
  • Total local-AI storage view
  • Basic menu-bar status

Recommended

Bonsai Pro

$29 $39

Early Access price · $39 regular

One purchase. Bonsai 1.x is yours. No subscription.

Get Bonsai Pro
  • Everything in Free
  • Full ranked recommendations, with “Why this model?” explanations
  • One-click Ollama install, with a post-install benchmark and predicted vs. actual comparison
  • Complete inventory across Ollama, LM Studio, and Hugging Face
  • Duplicate detection, stale-model insights, and safe cleanup with reclaimable-storage analysis
  • Full menu-bar monitoring with performance interpretation
  • Better-fit suggestions over time, with Pulse notifications when something meaningfully better fits

A few good questions

A little more
clarity.

The right model. The right space.

A little guidance.
A better fit.

Make local AI feel at home on your Mac. Get early access to Bonsai.

or take a closer look In development for macOS 14+.