The AI Bench
48 GB · 307 GB/s

What can the MacBook Pro (M5 Pro, 48 GB) run?

A laptop that runs 27–35B models. The sweet spot for developers on the move.

Best models for each task

Our top three for this machine. Chat and image tasks assume an 8K-token conversation; coding, documents and agents assume 32K.

Chat & writing20 of 25 run

  1. Qwen3.8 27BRuns wellQ8_0 · needs 31.6 GB at 8K · 6–10tok/sReadable
  2. Gemma 4 31BRuns wellQ8_0 · needs 35.7 GB at 8K · 5.5–9tok/sReadable
  3. Muse Glimmer 30BRuns wellQ8_0 · needs 32.1 GB at 8K · 6–10tok/sReadable

Coding21 of 26 run

  1. Qwen3.8 27BRuns greatUD-IQ3_S · needs 16.1 GB at 32K · 13–20tok/sSmooth
  2. Muse Glimmer 30BRuns greatQ4_K_M · needs 20.2 GB at 32K · 10–17tok/sSmooth
  3. Gemma 4 26B-A4BRuns wellQ8_0 · needs 29.9 GB at 32K · 25–55tok/sFast

Long documents15 of 20 run

  1. Qwen3.8 27BRuns greatUD-IQ4_XS · needs 18.4 GB at 32K · 11–18tok/sSmooth
  2. Gemma 4 26B-A4BRuns wellQ8_0 · needs 29.9 GB at 32K · 25–55tok/sFast
  3. Qwen3.6 35B-A3BRuns greatQ4_K_M · needs 24.9 GB at 32K · 40–85tok/sFast

Agents & tools21 of 26 run

  1. Qwen3.8 27BRuns greatUD-IQ3_S · needs 16.1 GB at 32K · 13–20tok/sSmooth
  2. Qwen3.6 35B-A3BRuns greatQ4_K_M · needs 24.9 GB at 32K · 40–85tok/sFast
  3. Ornith 1.5 35B-A3BRuns greatQ4_K_M · needs 23.4 GB at 32K · 40–85tok/sFast

Images in9 of 13 run

  1. Qwen3.8 27BRuns wellQ8_0 · needs 31.6 GB at 8K · 6–10tok/sReadable
  2. Muse Glimmer 30BRuns wellQ8_0 · needs 32.1 GB at 8K · 6–10tok/sReadable
  3. Qwen3.6 35B-A3BRuns greatQ4_K_M · needs 24.4 GB at 8K · 45–100tok/sFast

Make images6 of 6 run

  1. Stable Diffusion XL Base 1.0Runs greatFP16 · needs 7.9 GB
  2. FLUX.2 [klein] 4BRuns greatBF16 · needs 14.0 GB
  3. Z-Image-TurboRuns greatBF16 · needs 17.0 GB

Speak text7 of 7 run

  1. Qwen3-TTS 1.7BRuns greatBF16 · needs 5.5 GB
  2. Kokoro-82MRuns greatFP32 · needs 1.4 GB
  3. Chatterbox Multilingual V3Runs greatFull · needs 4.2 GB

Transcribe7 of 7 run

  1. Voxtral Mini 4B RealtimeRuns greatBF16 · needs 17.0 GB
  2. Whisper large-v3-turbo (via faster-whisper)Runs greatFP16 · needs 2.6 GB
  3. Nemotron 3.5 ASR Streaming 0.6BRuns greatFP32 · needs 3.5 GB

Biggest model it runs well

Kimi Linear 48B-A3B (Q4_K_M, 30.1 GB file) at about 45–100tok/sFast.

Should you buy it?

Buy it if

You need good local models on a laptop.

Skip it if

You stay at a desk — a Mac Studio has twice the bandwidth.

Watch out for

  • Long sessions on battery drain it fast.

Measured on this hardware

Real runs from public sources, not estimates. All measurements.

ModelVersionPromptSpeedSource
Qwen3.8 27BUD-Q4_K_XL51213.4 tok/sOurs 2026-10-10
Qwen3.8 27BUD-Q4_K_XL32K10.28 tok/sOurs 2026-10-10