The best local LLM for a MacBook Air M1 (8 GB) is Maple Preview 20B-A1B at 130 tok/s (estimated). With 8 GB of unified memory it runs 18 of the models in the LLMCheck index — from compact options up to 27B-class models. For everyday chat and coding, Maple Preview 20B-A1B is the sweet spot. Full ranking below.
Considering this Mac?
Mac mini M5 Pro on Amazon → · compare every Mac for local AI →
Ranked by LLMCheck suitability (capability balanced against speed on the M1). Click a model for its full benchmark and setup. Every speed carries its provenance mark: most are index estimates from the memory-bandwidth model, and a filled square is a published measurement — submit a real run →
| # | Model | Size | License | Speed | Capability |
|---|---|---|---|---|---|
| 1 | Maple Preview 20B-A1B | 20B | MIT | estimated130 tok/s | 20/50 |
| 2 | Bonsai 27B | 27B | Apache 2.0 | estimated3 tok/s | 33/50 |
| 3 | Nanbeige4.2-3B | 3B | Apache 2.0 | estimated31 tok/s | 20/50 |
| 4 | Qwen 3.5 9B | 9B | Apache 2.0 | estimated10 tok/s | 18/50 |
| 5 | Gemma 4 E4B | 4B | Apache 2.0 | estimated23 tok/s | 16/50 |
| 6 | Gemma 4 E2B | 2B | Apache 2.0 | estimated45 tok/s | 13/50 |
| 7 | LFM2.5-2.6B | 2.6B | LFM Open | estimated35 tok/s | 14/50 |
| 8 | DeepSeek R1 8B | 8B | MIT | estimated12 tok/s | 16/50 |
| 9 | Apertus 1.5 8B | 8B | Apache 2.0 | estimated11 tok/s | 16/50 |
| 10 | Phi-4 Mini | 3.8B | MIT | estimated24 tok/s | 14/50 |
| 11 | Qwen 3 8B | 8B | Apache 2.0 | estimated12 tok/s | 15/50 |
| 12 | Ministral 8B | 8B | Apache 2.0 | estimated12 tok/s | 14/50 |
Showing the top 12 of 18 models that fit in 8 GB. See the full leaderboard or all benchmarks.
The fastest way to get started is Ollama. Install it, then find the top pick for your Mac in its library and copy the exact tag it lists:
Find Maple Preview 20B-A1B in the Ollama library → · model card on Hugging Face →
Prefer a GUI? LM Studio gives you a one-click download and chat window. For step-by-step help see our Ollama install guide, or open the Maple Preview 20B-A1B on M1 benchmark page for exact settings.
The MacBook Air M1 (8 GB) tops out at Bonsai 27B. Newer Apple Silicon with more unified memory runs larger, smarter models much faster:
As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence our rankings.
Maple Preview 20B-A1B (20B, MIT) is the best all-round pick at 130 tok/s (estimated) on the M1. If you want maximum speed, Gemma 4 E2B hits 45 tok/s (estimated); for maximum capability, Bonsai 27B still fits in 8 GB.
About 18 of the 89 models in the LLMCheck leaderboard fit in 8 GB of unified memory, from compact models up to Bonsai 27B (27B).
Not comfortably. A 70B model in Q4 needs ~40–44 GB; with 8 GB you should stick to models up to ~4 GB, such as Maple Preview 20B-A1B. For 70B, look at a 48 GB+ Mac.
8 GB is enough for compact models like Gemma and Phi Mini, but tight for anything above ~8B. Because Apple Silicon uses unified memory, that figure is both your system RAM and your VRAM.