How much RAM does a Mac need to run local AI models?

16 GB is the floor for anything useful and gets you 7B to 12B models at 4-bit. 32 GB is the first tier we would call comfortable: it is where Qwen 3.8 27B at Q4_K_M fits, since that file is 17.1 GB of weights resident. 64 GB opens the 70B class at 4-bit. Below 16 GB you are running toys, because macOS and your browser take their share of the same pool.

Why does unified memory matter more than the chip name?

On Apple Silicon there is no separate VRAM. The GPU and the CPU share one pool of memory with the operating system, Safari, and everything else you have open. A model has to fit in what is left, so an M4 with 16 GB loads exactly the same models as an M2 with 16 GB. The chip generation changes how fast tokens come out. The memory size changes what you can load at all, which is the harder limit.

Can a Mac run Stable Diffusion or Flux well?

It can run them, but it is a poor fit compared with a rented NVIDIA card. Diffusion is a compute bound workload, the Mac tooling lags the CUDA tooling, and a single high resolution image that takes seconds on an H100 can take minutes on a laptop while the fans run. One image for fun is fine locally. A batch of forty variations is where a Mac stops being the right tool.

Can I train a LoRA on a Mac?

Realistically, no. Training is more memory hungry and far more compute hungry than inference, and the Mac training tooling is thin. LoRA training on the hosted side is included on the Hosted plan at 2 trainings a month, and is not available on a credit pack.

Is there a Mac build of the Locally Uncensored desktop app?

No. The free open source desktop app is Windows and Linux only, and there is no native Mac build. A Mac reader gets a local runner of their choice, such as LM Studio, Ollama or an MLX setup, plus the browser Studio at lu-labs.ai for the models and the image and video lanes the machine cannot handle.

What does the hosted side cost when my Mac runs out of room?

A 5 EUR credit pack is 165,000 credits with no subscription and credits never expire. Flux Schnell is 300 credits per image, Flux 2 Dev is 1,200 credits per image, and a 5 second LTX-2 clip is 8,000 credits, so that pack is roughly 550 quick images or 137 Flux 2 Dev images. The Hosted plan is 19 EUR a month with 900,000 credits. Full rates at lu-labs.ai/pricing.

What a Mac Can Actually Run Locally in 2026

Every Mac buying guide for local AI leads with the chip. M3, M4, Pro, Max, Ultra. That is the wrong number to look at. On Apple Silicon there is no separate VRAM: the GPU and the CPU share one pool of unified memory with macOS, your browser, your editor, and the twenty tabs you forgot about. A model has to fit in whatever is left over. So the spec that decides what you can load is the memory tier of the machine, and everything else decides how fast the tokens come out.

Practical version: an M4 with 16 GB loads exactly the same models as an M2 with 16 GB. It answers faster. It does not answer bigger.

The Tiers, By RAM

Rough working budget: assume macOS plus a browser eats 4 to 6 GB before you start, and that a model needs its weight file plus 1 to 2 GB of overhead for context. Everything below is 4-bit unless stated.

Unified memory	What fits	Verdict
8 GB	Small models only, in the 4B to 7B range at aggressive quants	Not a serious local AI machine. Use the browser
16 GB	7B to 12B comfortably, a 14B if you close everything else	The floor. Fine for a private assistant, drafts, summaries
24 GB	Adds the 20B class and a 27B dense model at a low quant	An awkward middle. Better than 16, short of the good stuff
32 GB	Qwen 3.8 27B at Q4_K_M, which is 17.1 GB of weights resident	The first tier we would actually recommend
48 to 64 GB	The 70B class at 4-bit, with room for long context	Serious. Mac Studio and high spec MacBook Pro territory
128 GB and up	Very large mixture of experts models at low quants	Possible, slow, and a lot of money for a niche

Note what is missing from that table: a 405B model. There is no configuration of laptop that runs one at usable quality, and no quant that changes the arithmetic.

The 32 GB Threshold, And Why We Are Blunt About It

If you are buying a Mac and local AI is part of the reason, 32 GB is the number. Below it you are permanently one model generation behind: the interesting open weight releases keep landing in the 27B to 35B range, and a 16 GB machine cannot hold them no matter how patient you are. At 32 GB, Qwen 3.8 27B fits at Q4_K_M with room for context, and a 0.9 GB companion file gives it image input. Our Qwen 3.8 27B setup guide has the quant table if you want the exact file sizes.

Apple charges a lot for that step. Pay it once, or plan on the hosted lane covering the gap. Those are the two honest options, and buying 16 GB while intending to run 35B models is not one of them.

If you are starting from zero on a Mac, LU Labs has a walkthrough on how to run AI models locally on your Mac, from picking a model size to the first chat.

The Half Nobody Mentions: Diffusion

Text generation on a Mac is genuinely good. Memory bandwidth on Apple Silicon is high and the MLX tooling has matured. Image and video generation is a different story, and this is where the marketing quietly stops.

Diffusion is compute bound rather than bandwidth bound, and it is the workload NVIDIA hardware was tuned for over a decade. The Mac tooling also lags: most new samplers, ControlNets and video pipelines land on CUDA first and arrive on Apple Silicon months later, if at all. In practice a single high resolution image that finishes in seconds on rented hardware can take minutes on a laptop with the fans at full tilt, and a video clip is worse by an order of magnitude. If you want the local diffusion route anyway, our Flux local guide covers what it costs in hardware.

This is where the cloud earns its keep, and we would rather say that plainly than sell a Mac workflow that does not exist. The hosted lanes at LU Labs run on managed NVIDIA H100, A100 and B200 class GPUs, with 10 image models, editing and upscale, and 5 video models producing 5 or 8 second clips. The hardware is shared, so at busy hours there is a short queue.

Where the Line Falls in Practice

Task	Local Mac	Hosted
Short chat, drafting, summaries	Yes, and it should stay local	Only if you want a bigger model
Long agent loops with tool calls	Painful, and it heats the machine for an hour	Yes
A trillion parameter model	No	Yes, Kimi K3
One image, for fun	Workable, expect minutes	Yes, and it is seconds
A batch of forty variations	No, this is an evening of waiting	Yes
A 5 second video clip	Not really	Yes, LTX-2 and four other models
LoRA training	No	2 a month on the Hosted plan, none on a credit pack

What the Other Side Costs

Concrete anchors, because "cloud is expensive" is usually said without numbers. Flux Schnell is 300 credits per image. Flux 2 Dev is 1,200 credits per image. A 5 second LTX-2 clip is 8,000 credits. A 5 EUR pack is 165,000 credits, needs no subscription, and the credits never expire, which works out to roughly 550 quick images or 137 Flux 2 Dev images. The Hosted plan is 19 EUR a month with 900,000 credits, about 3,000 quick images or 750 Flux 2 Dev images, plus the LoRA training slots. Per model rates are at https://lu-labs.ai/pricing.

Compare that against the Apple price of stepping from 16 GB to 32 GB and it is not a close call for image work.

The Part We Will Not Dress Up

Our free open source desktop app is Windows and Linux only. There is no native Mac build. So a Mac reader ends up with two pieces: a local runner of your choice for the offline half, and the browser Studio for everything the machine cannot hold. The LM Studio alternative for Mac guide walks that setup step by step, including the API key path so the hosted models answer from your terminal too.

If you want to compare the local runners themselves first, the best local AI apps of 2026 roundup covers which ones have decent Apple Silicon support.

Locally Uncensored is AGPL-3.0 licensed and built by PurpleDoubleD. Corrections welcome on GitHub Discussions or in the Discord.