An LM Studio alternative for Mac, in the browser.

The honest starting point

There is no native Mac build of the LU Labs desktop app. It ships for Windows and Linux today. So this page is not offering you a second local runner to install next to LM Studio. What it offers is the other half of the job: the models that never fit on a laptop, plus the image and video work LM Studio does not do, reachable from Safari.

LM Studio and LU Labs Cloud, side by side

What LM Studio is good at

• It runs open-weight chat models on Apple silicon, and it is free.
• Everything stays on your disk. No account, no connection, no bill.
• It works on a plane, and it keeps working when a service goes down.

Where it runs out of room

• The model has to fit in your Mac's memory. A 16 GB machine tops out around the small models, and buying more RAM means buying a new Mac.
• It is a text tool. There is no image studio, no video studio, no LoRA training.
• Large models are slow or simply out of reach on a laptop.

What Cloud adds

• 15 chat models on the entry plan, including models far larger than a laptop can hold, all of them able to call tools in Agent and Code mode.
• 10 image models and 5 video models, with editing, background removal, upscale and character training.
• Nothing to install or update. No GGUF files on your SSD, no quantization to pick, no runtime to keep current.
• A personal API key on an OpenAI-compatible endpoint, so your editor and your scripts can use the same account.

How to start in three steps

1. Take the Hosted plan at checkout, or a credit pack if you would rather pay once. Card details go to Stripe and the payment is secured with 3D Secure.
2. The Studio opens in your browser on your Mac. Nothing downloads.
3. Pick a model and start, or switch to the Image, Video or Code tab.

What it costs

The honest limits

Cloud is not local, and that is the whole trade. Your prompt leaves your Mac and is processed on our fleet, so if the point of LM Studio for you is that nothing ever leaves the machine, keep LM Studio. What we can promise instead: we never train on your data and never sell it, we host open-weight models only, and account data lives in the EU where you can delete it. The rest of the fine print: shared hardware means a request can queue for a few seconds at peak, credits are one pool for text, images and video, the model catalog is curated rather than open ended, and a connection is required. Many people run both, LM Studio for the offline everyday and Cloud for the jobs a laptop cannot hold.

Questions