Used Mac & MacBook fair prices

Entry local AI Mac. Fine for 7B–8B class models via Ollama/MLX. Not ideal if you want 30B+ resident — step up RAM.

24GB shared by macOS, applications and inference. Check the exact model artifact and context before choosing this configuration.

48GB shared memory offers more capacity than the 24GB configuration. Allow for macOS and applications, then compare measured runtime performance.

Used-market favorite: 32GB unified memory often cheaper than new high-RAM M4 configs. Excellent value for local LLMs if you find a clean unit.

16GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities; verify the full configuration and seller return terms before buying.

24GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities; verify the full configuration and seller return terms before buying.

64GB opens larger models and longer context than any 24GB GPU. Used Studios with high RAM are the secondary-market play for serious local AI without NVIDIA.

64GB unified memory and an Ultra chip. Verify the listed configuration and available memory; use a matching workload benchmark to assess speed.

128GB unified memory for workloads needing more capacity. Allow for system and runtime overhead; check the exact artifact and context.

Current-gen Studio Ultra. Buy by RAM tier first (96GB+), not just chip name. New/refurb Amazon + used eBay both apply.

36GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities; verify the full configuration and seller return terms before buying.

96GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities; verify the full configuration and seller return terms before buying.

32GB unified memory. Confirm model and runtime requirements, then inspect battery health, keyboard and seller return terms.

64GB portable configuration. Check available memory and a representative sustained workload; neither model fit nor battery runtime follows from capacity alone.

32GB unified memory. Compare current prices and matched workload benchmarks with other Max configurations; memory is shared with macOS.

16GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities and screen sizes; verify the full configuration and seller return terms before buying.

24GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities and screen sizes; verify the full configuration and seller return terms before buying.

36GB unified memory is shared by the operating system, applications, and model weights. Check the memory requirements of your exact model and context length. Samples group storage capacities and screen sizes; verify the full configuration and seller return terms before buying.

8GB shared memory leaves limited space for local inference. Small models may be usable; test the workload before buying a machine specifically for AI.

16GB shared memory in a fanless laptop. Check artifact size, context and sustained workload behavior before purchasing for local inference.

Recent Air with 16GB standard on later configs. Slightly faster than M2 for small local models; buy on price against the M2 16GB, not on the chip bump.

16GB is shared by macOS, applications, and model memory. Compare 24GB or 32GB configurations if local AI is the main purchase reason. Samples group screen sizes and storage capacities; check both on the listing.