Run on your Mac

Run Outlier Lite on MacBook Pro M1 Pro

Last updated 2026-06-18 · Outlier v1.11.469

What does Outlier Lite run like on MacBook Pro M1 Pro?

The pairing under examination is Outlier Lite (5.04 GB MLX 4-bit, Qwen3.5-9B base, 32K default context, 12 GB unified memory minimum) on a MacBook Pro M1 Pro (Apple M1 Pro, 8-10 CPU cores, 14-16 GPU cores, 16|32 GB unified memory, 200 GB/s, 2021). Outlier loads this tier as one mlx_lm process inside the bundled FastAPI sidecar. Apple Silicon decode on dense 4-bit weights is bandwidth-bound, so the ceiling on this exact MacBook Pro M1 Pro scales as the ratio 200 / 800 = 0.25× the published M1 Ultra number.

How much memory does the model take during generation?

Outlier’s Outlier Lite has a measured peak generation footprint of about 5.5 GB. The MacBook Pro M1 Pro base configuration ships with 16|32 GB unified, and the rule of thumb is to leave roughly 4 GB for the OS and one open browser tab. On MacBook Pro M1 Pro the Outlier Lite bandwidth-scaled estimate is ~13.3 tok/s [estimated from family-bandwidth ratio]; derivation: 53.4 tok/s on M1 Ultra × (200 / 800) bus ratio for the Apple M1 Pro. Treat as a first-order projection — the Outlier Lite number on this 2021 machine has not been formally measured.

What is the install path on MacBook Pro M1 Pro for Outlier Lite?

The 2021 MacBook Pro M1 Pro ships with 16|32 GB of unified memory, so headroom for the 12 GB Outlier Lite requirement is 4 GB on the base SKU. Step-by-step install instructions live on the install guide; the part that varies for this Apple M1 Pro machine is that the 5.04 GB pull from Outlier-Ai/Outlier-Lite-9B-MLX-4bit over HTTPS lands in ~/Library/Application Support/Outlier/models/lite/ and download time is bandwidth-limited by the network, not by the 8-10-core Apple M1 Pro CPU.

What context window can I use on a 16 GB Mac?

The Outlier Lite tier defaults to 32K context and caps at 256K. KV cache scales linearly with context length on dense models, so longer contexts trade headroom for capacity. On a 16|32 GB MacBook Pro M1 Pro, the default context is the safe starting point.

What works well on this pairing, and what is still rough?

• The MacBook Pro M1 Pro (Apple M1 Pro, 200 GB/s) is bandwidth-bound on dense 4-bit decode; the throughput ceiling for Outlier Lite on this exact machine is approximately 200 divided by 5.04 GB tokens per second.
• The 16 GB base SKU of the MacBook Pro M1 Pro swaps aggressively under the Outlier Lite tier; close Chrome and other RAM-heavy applications before generation.
• The Apple M1 Pro memory bus runs slower than M2 and later silicon in the same class; the Outlier Lite tier numbers are correspondingly lower than the M4 equivalent.
• Cold boot of Outlier on a 8-10-core Apple M1 Pro is roughly 50 seconds while the bundled Python.framework warms; this is documented on the v1.9 backlog under the boot-time investigation thread.

What is the unique number for Outlier Lite on MacBook Pro M1 Pro?

5.04 GB on disk against 16|32 GB unified memory means the weights alone consume about 32% of the base SKU’s RAM on the MacBook Pro M1 Pro.

Should I pick a different tier?

For lighter-weight work on this Apple M1 Pro, the Nano tier is the next step down and runs visibly faster. For heavier work, the Quick or Core tier is the next step up; check that the 16|32 GB SKU you have can fit its weights before installing. The unified-memory explainer works the bandwidth math out, and the bandwidth ratio for MacBook Pro M1 Pro is 0.25× M1 Ultra reference.

Download Outlier for Mac

Requires Apple Silicon (M1, M2, M3, or M4) — Intel Macs are not supported. macOS 12+.