The best local LLM for a MacBook Pro M5 Pro (48 GB) is LFM2.5-24B-A2B at 114 tok/s (community run). Its 48 GB fits 71 of the 130 models in the LLMCheck index, up to 80B-class. For everyday chat and coding, LFM2.5-24B-A2B is the sweet spot.
Considering this Mac?
MacBook Pro M5 Pro (48 GB): build-to-order from Apple · MacBook Pro M5 Pro (24 GB) on Amazon → · not sure? the Mac Advisor picks a config →
Ranked by LLMCheck suitability (capability balanced against speed on the M5 Pro); 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 →
| # | Model | Size | License | Speed | 64K context | Capability |
|---|---|---|---|---|---|---|
| 1 | LFM2.5-24B-A2B | 24B MoE | LFM Open | community114 tok/s | agent-ready | 35/50 |
| 2 | Qwen3.8-27B | 27.8B | Apache 2.0 | sourced17 tok/s | agent-ready | 46/50 |
| 3 | Recursive Flint | 39.8B MoE | Apache 2.0 | community92 tok/s | agent-ready | 37/50 |
| 4 | KAT-Coder-V2.5 | 35B MoE | Apache 2.0 | estimated72 tok/s | agent-ready | 39/50 |
| 5 | Qwen 3.6-35B-A3B | 35B MoE | Apache 2.0 | community79 tok/s | agent-ready | 38/50 |
| 6 | Nanbeige4.3-33B slow | 33.1B | Apache 2.0 | estimated12 tok/s | agent-ready | 46/50 |
| 7 | Qwen 3.6-27B | 27B | Apache 2.0 | estimated15 tok/s | agent-ready | 44/50 |
| 8 | Nemotron 3.5 Lightning | 30B MoE | OpenMDW | community72 tok/s | agent-ready | 36/50 |
| 9 | North Mini Code 1.0 | 30.5B MoE | Apache 2.0 | community70 tok/s | agent-ready | 36/50 |
| 10 | Gemma 4 26B-A4B | 26B MoE | Apache 2.0 | estimated78 tok/s | agent-ready | 35/50 |
| 11 | Raon-1 mini preview | 28.8B MoE | Apache 2.0 | community86 tok/s | agent-ready | 34/50 |
| 12 | GLM-4.7-Flash | 31B MoE | MIT | estimated73 tok/s | agent-ready | 35/50 |
Showing the top 12 of 71 models that fit in 48 GB. See the full leaderboard or all benchmarks.
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 42 GB this Mac’s GPU can use with the limit raised (57 of the 66 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 37.4 GB of this Mac’s 48 GB by default (macOS 26 and later, the only macOS it runs). Of the 71 models that fit, 7 need more than that at a short prompt: 3 need a raised limit and are tagged raise GPU limit (Nemotron 3.5 Bolt 74B-A9B and Llama 3.3 70B only past the safe setting below, as measured runs on this much memory show); 4 need more than even the raised limit and are tagged over GPU limit. sudo sysctl iogpu.wired_limit_mb=43008 raises it to 42 GB, leaving 6 GB for macOS, until the Mac restarts (sudo sysctl iogpu.wired_limit_mb=0 restores the default). How the limit is read →
The fastest way to get started is Ollama. Install it, then pull the model by its tag (checked on ollama.com, 2026-11-14):
ollama run lfm2.5:24b · 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 LFM2.5-24B-A2B on M5 Pro page for its figures and setup.
Twice as fast: no Mac in the index runs LFM2.5-24B-A2B at twice this Mac’s community114 tok/s; the fastest figure is the Mac Studio M5 Max (48 GB) at sourced184 tok/s, $3,099 new (Apple list price, checked 2026-11-14): build-to-order from Apple (Amazon does not stock this configuration).
For less: the cheapest Mac that still runs LFM2.5-24B-A2B at 15+ tok/s is the MacBook Air M2 (24 GB) at estimated~52 tok/s, about $850 refurbished or used: search Amazon Renewed listings →. New: the Mac mini M6 (24 GB) at estimated~79 tok/s, $1,099 new (Apple list price, checked 2026-11-14): Mac mini M6 (24GB) 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.
The MacBook Pro M5 Pro (48 GB) runs 71 of the models in the LLMCheck index, led by LFM2.5-24B-A2B. This 48 GB configuration is build-to-order from Apple: Amazon does not stock it (it carries the standard 24 GB MacBook Pro M5 Pro, with less memory).
As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence our rankings.
LLMCheck. (2026). Best Local LLMs for the MacBook Pro M5 Pro (48 GB). llmcheck.net. https://llmcheck.net/best-llm/macbook-pro-m5-pro-48gb/ (updated ). CC BY 4.0.
LFM2.5-24B-A2B (24B MoE, LFM Open) is the best all-round pick at 114 tok/s (community run) on the MacBook Pro M5 Pro. If you want maximum speed, Maple Preview 20B-A1B hits 387 tok/s (published figure); for maximum capability, Qwen3.8-27B still fits in 48 GB.
About 71 of the 130 models in the LLMCheck index fit in 48 GB of unified memory, from compact models up to Qwen3-Coder-Next (80B MoE).
Just. A dense 70B such as Llama 3.3 70B needs about 44 GB at 4-bit, which leaves little of the 48 GB for context and other apps; the index puts it at ~6 tok/s (estimated) here.
48 GB is enough for dense models up to about 72B at 4-bit (the largest in the index that fits is Qwen 2.5 72B), and for mixture-of-experts models such as Nemotron 3.5 Bolt 74B-A9B (~31 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.