Outlier
AI & ML interests
None defined yet.
Recent Activity
Outlier
Outlier runs AI models on your own Mac, with no cloud and no subscription. Nothing you type leaves the Mac unless you turn on web search. Free Nano and Lite; Pro is $49 once ($24.50 through 30 November). Mac app 1.11.999 (6 November 2026), also brew install --cask outlier-host/tap/outlier.
- Free: Outlier Nano and Outlier Lite.
- Pro: $49 once ($24.50 through 30 November), on up to 3 Macs, with a 30-day refund.
- Apple silicon, macOS 15 or later.
outlier.host · Download for Mac
The models in the app (November 2026)
| Model | Plan | Disk | Min RAM |
|---|---|---|---|
| Outlier Nano (Qwen3.5-4B, MLX 4-bit) | Free | 2.37 GB | 6 GB |
| Outlier Lite (Qwen3.5-9B, MLX 4-bit) | Free | 5.04 GB | 12 GB |
| Outlier Quick (Gemma-4-26B-A4B) | Pro | 15.61 GB | 16 GB |
| Outlier Core (Qwen3.6-27B) | Pro | 15.13 GB | 24 GB |
| Outlier Plus (Qwen3.5-397B-A17B, streams from the SSD) | Pro | 209 GB | 64 GB |
| Outlier Vision 3.8 (Qwen3.8-27B, reads images) | Pro | 15.5 GB | 24 GB |
The benchmarks are at outlier.host/data.
Patents
Three U.S. provisional patent applications were filed in April 2026. They cover ternary mixture-of-experts inference, paged expert routing on consumer hardware, and zero-delta initialization with residual error correction.
Research artifacts
The V3.2 models (10B, 40B, 70B and 150B) on this page are research models from the training work that went into v1.8. They aren't the app. The app is the Mac download at outlier.host. The earlier Outlier-10B, Outlier-10B-V2 and Outlier-40B repos stay up so old links keep working.
Built by
Matt Kerr · Kerr & Company LLC · Grand Rapids, MI · matt@outlier.host
-
Outlier-Ai/Outlier-70B-V3.3
Text Generation • Updated • 576 • 1 -
Outlier-Ai/Outlier-Lite-7B-MLX-4bit
Text Generation • 8B • Updated • 54 -
Outlier-Ai/Qwen3-Coder-30B-A3B-Instruct-MLX-4bit
Text Generation • 31B • Updated • 977 • 1 -
Outlier-Ai/Outlier-Max-32B-GGUF
Text Generation • 33B • Updated • 200 • 1
-
Outlier-Ai/Outlier-70B-V3.3
Text Generation • Updated • 576 • 1 -
Outlier-Ai/Outlier-Lite-7B-MLX-4bit
Text Generation • 8B • Updated • 54 -
Outlier-Ai/Qwen3-Coder-30B-A3B-Instruct-MLX-4bit
Text Generation • 31B • Updated • 977 • 1 -
Outlier-Ai/Outlier-Max-32B-GGUF
Text Generation • 33B • Updated • 200 • 1