OwnRig

Gemma 2 27B Instruct on Apple M4 Max (36GB Unified)

Apple M4 Max (36GB Unified) handles Gemma 2 27B Instruct well at 15 tok/s at Q5_K_M. A solid choice for this model.

Model Size

27.23B

Device VRAM

36 GB

Bandwidth

546 GB/s

Quantizations Tested

1

Performance by Quantization

Each row shows Gemma 2 27B Instruct performance at a different quality level on Apple M4 Max (36GB Unified).

QuantizationSpeedTTFTFits in VRAMRatingConfidence
Q5_K_M15 tok/s600ms✓ YesGoodestimated

Notes

Q5_K_M

Q5 at 18.5GB in 36GB. Good quality with headroom for a small concurrent model.

About Gemma 2 27B Instruct

Gemma 2 27B Instruct (27.23B) is a chat, coding, reasoning, multi-purpose model. Google's 27B model with excellent knowledge distillation. Strong reasoning and coding capabilities at a size that fits on a single 24GB GPU at Q4. Limited to 8K context.

View all Gemma 2 27B Instruct hardware options →

About Apple M4 Max (36GB Unified)

Apple M4 Max (36GB Unified) has 36 GB at 546 GB/s. Available in MacBook Pro 16", Mac Studio.

See all models Apple M4 Max (36GB Unified) can run →

Source: MLX performance estimates (2026-01-15)

Data last updated: 2026-03-01