Best Local LLMs for the MacBook Air M5 (16 GB)

The best local LLM for a MacBook Air M5 (16 GB) is Gemma 4 12B at ~17 tok/s (estimated), the top-ranked model that reaches 15 tok/s here. Its 16 GB fits 38 of the 130 models in the LLMCheck index, up to 24B-class. The most capable, Ternary Bonsai 2 27B, runs at 12 tok/s (community run): slow for chat.

Unified memory
16
GB
Mem. bandwidth
153
GB/s
Models that fit
38
of 130
Top speed
estimated127
tok/s

Considering this Mac?
MacBook Air M5 (16GB) on Amazon → · not sure? the Mac Advisor picks a config →

Top 3 picks for the MacBook Air M5 (16 GB)

⭐ Best overall
12B · Apache 2.0 · cap 30/50 · agent-ready 64K
estimated17 tok/s
⚡ Fastest
7.9B MoE · MIT · cap 15/50 · agent-ready 64K
estimated127 tok/s
🧠 Most capable (slow)
27.36B · Apache 2.0 · cap 37/50
community12 tok/s

Every model ranked for a MacBook Air M5 (16 GB)

Ranked by LLMCheck suitability (capability balanced against speed on the M5); the top pick is the best-ranked model that reaches 15 tok/s here, has a standard Mac runtime and fits the GPU memory limit (raised if need be), and slower models are tagged slow. Click a model for its full figures and setup. Every speed carries its provenance mark: most are index estimates from the memory-bandwidth model, a filled square is a published figure and a hollow circle a community run — submit a real run →

#ModelSizeLicenseSpeed64K contextCapability
1Gemma 4 12B12BApache 2.0estimated17 tok/sagent-ready30/50
2Ternary Bonsai 2 27B slow27.36BApache 2.0community12 tok/sneeds ~14 GB37/50
3Bonsai 27B27BApache 2.0—agent-ready33/50
4LFM3-1.7B1.73BLFM Opencommunity117 tok/sagent-ready17/50
5MiniCPM5-2B2.52BApache 2.0estimated82 tok/sagent-ready21/50
6Ling-3.0-tiny7.9B MoEMITestimated127 tok/sagent-ready15/50
7Nanbeige4.2-3B3BApache 2.0estimated64 tok/sagent-ready20/50
8GPT-oss 20B raise GPU limit21B MoEApache 2.0estimated41 tok/sneeds ~15 GB22/50
9K2-Horizon-3.7B5.1BApache 2.0estimated39 tok/sneeds ~14 GB22/50
10Mistral Small 3.2 24B slow over GPU limit24BApache 2.0estimated8 tok/sneeds ~25 GB25/50
11Devstral Small 24B slow over GPU limit24BApache 2.0estimated8 tok/sneeds ~25 GB24/50
12Spark-X2.6-4B4.11BApache 2.0estimated40 tok/sagent-ready20/50
Provenance key
sourced
Published figure — vendor or press review
estimated
community
Linked individual run — blog, forum or submission

Showing the top 12 of 38 models that fit in 16 GB. See the full leaderboard or all benchmarks.

Estimated figures come from the bandwidth formula; against 163 measured runs its typical error is ±9% for dense models and ±15% for small MoEs (≤40B total). How the error is computed →

Context: “fits” means the weights at a short prompt. Coding agents (Claude Code, OpenCode, Codex CLI) hold 64K tokens or more, and every token adds KV cache: agent-ready marks a model whose weights plus a 64K-token 16-bit KV cache still fit in the 12 GB this Mac’s GPU can use with the limit raised (20 of the 34 models here with a KV figure). That is an estimate from each model’s config.json, not a measurement; a q8 KV cache halves the cache part, and — means the index has no KV figure for the model yet. The GPU memory limit below applies on top.

GPU memory limit: macOS lets the GPU use about 11.8 GB of this Mac’s 16 GB by default (macOS 26 and later, the only macOS it runs). Of the 38 models that fit, 3 need more than that at a short prompt: 1 needs a raised limit and is tagged raise GPU limit (GPT-oss 20B only past the safe setting below, as measured runs on this much memory show); 2 need more than even the raised limit and are tagged over GPU limit. sudo sysctl iogpu.wired_limit_mb=12288 raises it to 12 GB, leaving 4 GB for macOS, until the Mac restarts (sudo sysctl iogpu.wired_limit_mb=0 restores the default). How the limit is read →

Quick start: run Gemma 4 12B on your MacBook Air M5

The fastest way to get started is Ollama. Install it, then pull the model by its tag (checked on ollama.com, 2026-11-14):

brew install ollama

ollama run gemma4:12b · model card →

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 Gemma 4 12B on M5 page for its figures and setup.

Same Mac with more memory, and the upgrade path

Upgrade or save: Gemma 4 12B on other Macs

Same pick, other Macs

Twice as fast: the cheapest Mac in the index that runs Gemma 4 12B at 2× or more of this Mac’s estimated~17 tok/s is the Mac Studio M2 Max (32 GB) at estimated~43 tok/s, about $1,300 refurbished or used: search Amazon Renewed listings →. New: the Mac mini M5 Pro (24 GB) at estimated~34 tok/s, $1,699 new (Apple list price, checked 2026-11-14): Mac mini M5 Pro (24GB) on Amazon →.

For less: the cheapest Mac that still runs Gemma 4 12B at 15+ tok/s is the Mac mini M2 Pro (16 GB) at estimated~22 tok/s, about $700 refurbished or used: search Amazon Renewed listings →. New: the Mac mini M6 (16 GB) at estimated~17 tok/s, $899 new (Apple list price, checked 2026-11-14): Mac mini M6 (16GB) on Amazon →.

Speeds are each Mac’s own figure from the index, with its provenance mark. Prices come from data/macs.json: Apple’s list price for a new Mac (dated), an approximate street price for a used or refurbished one. Amazon stock checked 2026-11-18. Amazon links are affiliate links: as an Amazon Associate, LLMCheck earns from qualifying purchases.

🛒 Get a MacBook Air M5 for local AI

The MacBook Air M5 (16 GB) runs 38 of the models in the LLMCheck index, led by Gemma 4 12B. Grab one and start running LLMs offline today:

As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence our rankings.

Cite this page

LLMCheck. (2026). Best Local LLMs for the MacBook Air M5 (16 GB). llmcheck.net. https://llmcheck.net/best-llm/macbook-air-m5-16gb/ (updated ). CC BY 4.0.

How to cite LLMCheck →Data: CC BY 4.0Markdown version

FAQ: local LLMs on the MacBook Air M5

What is the best local LLM for a MacBook Air M5 (16 GB)?

Gemma 4 12B (12B, Apache 2.0) is the best all-round pick at ~17 tok/s (estimated) on the MacBook Air M5. If you want maximum speed, Ling-3.0-tiny hits ~127 tok/s (estimated); Ternary Bonsai 2 27B is more capable but runs at only 12 tok/s (community run) here.

How many models can a MacBook Air M5 with 16 GB run?

About 38 of the 130 models in the LLMCheck index fit in 16 GB of unified memory, from compact models up to Mistral Small 3.2 24B (24B).

Can a MacBook Air M5 run a 70B model?

No. A dense 70B such as Llama 3.3 70B needs about 44 GB at 4-bit; with 16 GB, stick to models like Gemma 4 12B. For a 70B, look at a 64 GB+ Mac.

Is 16 GB of RAM enough to run LLMs locally?

16 GB is enough for dense models up to about 24B at 4-bit (the largest in the index that fits is Mistral Small 3.2 24B), and for mixture-of-experts models such as GPT-oss 20B (~41 tok/s (estimated)). As a rule of thumb a 4-bit model needs about 0.57 GB per billion parameters plus ~2 GB, and macOS needs its own share of memory too. Because Apple Silicon uses unified memory, the same pool serves as system RAM and GPU memory.

Related