Yes: Nemotron 3.5 Bolt E4B runs on 53 of 54 Macs in the LLMCheck index. It needs about 6 GB at a 4K context, 8 GB at 64K. The smallest and cheapest current Mac that runs it at 15+ tok/s is the Mac mini M6 16 GB ($899; ~42 tok/s est.). Best measured: 46.4 tok/s on an M5 Pro (community).
One row per Mac configuration, grouped by product line. The build is the one each speed is for; the context column is the longest prompt that fits inside macOS’s default GPU memory limit with a 16-bit KV cache (“raised” where the limit has to be raised first; “full” when the whole window fits). Speeds are decode tok/s from the LLMCheck index, each with its provenance mark. For dense models, the formula’s typical error is about 9% (the median gap to 75 measured plain-decode runs; it tends to undershoot (median −6%), and an estimate scaled from the model’s own measured run is usually closer; method).
| Mac | Verdict | Build | Speed | Max context | Model page |
|---|---|---|---|---|---|
| MacBook Air | |||||
| MacBook Air M5 16 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| MacBook Air M5 24 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| MacBook Air M5 32 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| MacBook Air M1 8 GB refurb | Over GPU limit why ↓ | Q4_K_M | estimated19 tok/s | — | — |
| MacBook Air M4 16 GB used | Runs | Q4_K_M | estimated33 tok/s | 64K (full) | M4 page |
| MacBook Air M2 16 GB refurb | Runs | Q4_K_M | estimated28 tok/s | 64K (full) | — |
| MacBook Air M3 16 GB refurb | Runs | Q4_K_M | estimated28 tok/s | 64K (full) | — |
| MacBook Air M1 16 GB refurb | Runs | Q4_K_M | estimated19 tok/s | 64K (full) | — |
| MacBook Air M4 24 GB used | Runs | Q4_K_M | estimated33 tok/s | 64K (full) | M4 page |
| MacBook Air M2 24 GB refurb | Runs | Q4_K_M | estimated28 tok/s | 64K (full) | — |
| MacBook Air M3 24 GB refurb | Runs | Q4_K_M | estimated28 tok/s | 64K (full) | — |
| MacBook Pro | |||||
| MacBook Pro M6 16 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| MacBook Pro M5 Pro 24 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| MacBook Pro M6 24 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| MacBook Pro M6 32 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| MacBook Pro M5 Pro 48 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| MacBook Pro M5 Max 64 GB | Runs | Q4_K_M | estimated161 tok/s | 64K (full) | M5 Max page |
| MacBook Pro M5 Pro 64 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| MacBook Pro M5 Max 128 GB | Runs | Q4_K_M | estimated161 tok/s | 64K (full) | M5 Max page |
| MacBook Pro M3 Max 64 GB refurb | Runs | Q4_K_M | estimated100 tok/s | 64K (full) | M3 Max page |
| MacBook Pro M3 Max 128 GB refurb | Runs | Q4_K_M | estimated100 tok/s | 64K (full) | M3 Max page |
| Mac mini | |||||
| Mac mini M6 16 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| Mac mini M5 Pro 24 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| Mac mini M6 24 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| Mac mini M6 32 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| Mac mini M5 Pro 48 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| Mac mini M5 Pro 64 GB | Runs | Q4_K_M | estimated79 tok/s | 64K (full) | M5 Pro page |
| Mac mini M2 Pro 16 GB refurb | Runs | Q4_K_M | estimated54 tok/s | 64K (full) | — |
| Mac mini M4 16 GB used | Runs | Q4_K_M | estimated33 tok/s | 64K (full) | M4 page |
| Mac mini M4 Pro 24 GB used | Runs | Q4_K_M | estimated72 tok/s | 64K (full) | — |
| Mac mini M4 24 GB used | Runs | Q4_K_M | estimated33 tok/s | 64K (full) | M4 page |
| Mac mini M2 Pro 32 GB refurb | Runs | Q4_K_M | estimated54 tok/s | 64K (full) | — |
| Mac mini M4 Pro 48 GB used | Runs | Q4_K_M | estimated72 tok/s | 64K (full) | — |
| Mac mini M4 Pro 64 GB used | Runs | Q4_K_M | estimated72 tok/s | 64K (full) | — |
| iMac | |||||
| iMac M6 16 GB | Runs | Q4_K_M | estimated42 tok/s | 64K (full) | — |
| iMac M6 24 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| iMac M6 32 GB | Runs | Q4_K_M | estimated46 tok/s | 64K (full) | — |
| Mac Studio | |||||
| Mac Studio M5 Max 36 GB | Runs | Q4_K_M | estimated113 tok/s | 64K (full) | M5 Max page |
| Mac Studio M5 Max 48 GB | Runs | Q4_K_M | estimated161 tok/s | 64K (full) | M5 Max page |
| Mac Studio M5 Max 64 GB | Runs | Q4_K_M | estimated161 tok/s | 64K (full) | M5 Max page |
| Mac Studio M5 Ultra 96 GB | Runs | Q4_K_M | estimated261 tok/s | 64K (full) | — |
| Mac Studio M5 Max 128 GB | Runs | Q4_K_M | estimated161 tok/s | 64K (full) | M5 Max page |
| Mac Studio M5 Ultra 256 GB | Runs | Q4_K_M | estimated261 tok/s | 64K (full) | — |
| Mac Studio M5 Ultra 512 GB | Runs | Q4_K_M | estimated261 tok/s | 64K (full) | — |
| Mac Studio M2 Max 32 GB refurb | Runs | Q4_K_M | estimated100 tok/s | 64K (full) | — |
| Mac Studio M1 Ultra 64 GB refurb | Runs | Q4_K_M | estimated196 tok/s | 64K (full) | — |
| Mac Studio M4 Max 64 GB used | Runs | Q4_K_M | estimated147 tok/s | 64K (full) | — |
| Mac Studio M2 Max 64 GB refurb | Runs | Q4_K_M | estimated100 tok/s | 64K (full) | — |
| Mac Studio M1 Ultra 128 GB refurb | Runs | Q4_K_M | estimated196 tok/s | 64K (full) | — |
| Mac Studio M2 Ultra 128 GB refurb | Runs | Q4_K_M | estimated196 tok/s | 64K (full) | — |
| Mac Studio M4 Max 128 GB used | Runs | Q4_K_M | estimated147 tok/s | 64K (full) | — |
| Mac Studio M2 Ultra 192 GB refurb | Runs | Q4_K_M | estimated196 tok/s | 64K (full) | — |
| Mac Studio M3 Ultra 256 GB used | Runs | Q4_K_M | estimated199 tok/s | 64K (full) | — |
| Mac Studio M3 Ultra 512 GB refurb | Runs | Q4_K_M | estimated199 tok/s | 64K (full) | — |
Context memory is computed from its config.json (config.json of nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16, the base checkpoint the 4GB quant names (that repo has none). 4 full layers (2 kv heads x 512) + 20 sliding layers (2 x 256, window 512); the last 18 layers share earlier layers' KV and store none): about 1.1 GB of 16-bit KV cache at 64K tokens. A q8 cache (llama.cpp -ctk q8_0 -ctv q8_0, Ollama OLLAMA_KV_CACHE_TYPE=q8_0) halves that part. Runtimes also grow a compute buffer with context, so a fit within about 1 GB of the limit is optimistic.
macOS lets the GPU wire only part of unified memory by default (Metal’s recommended working set; M5 and M6
Macs run macOS 26 or later, whose default is higher), and Nemotron 3.5 Bolt E4B needs about 6 GB. On the Macs marked
Raise GPU limit it fits in RAM but not inside that default. The default
is a performance threshold, not a hard wall: past it, the runtime moves part of it to the CPU or shortens the context (llama.cpp, Ollama), runs it slowly (MLX), or stops with an out-of-memory error. One command raises the limit until the next restart
(sudo sysctl iogpu.wired_limit_mb=0 restores the default sooner); it leaves macOS a reserve, and other
apps get less memory while the model is loaded. On the Macs marked Over GPU limit even that raise falls short: the model fits the Mac’s RAM on paper, but a smaller quant or a bigger Mac is the realistic route.
Weights are 4B parameters at each format’s effective size (the index’s GB-per-billion table), plus about 2 GB for the runtime and a short context; longer contexts add the KV cache computed from the model’s config.json at 16 bits. The first row is the catalog’s own figure (about 6 GB at a 4K context), the one every verdict on this page uses. “Smallest Mac” is the smallest memory size among the Macs Apple sells new that holds it at a 4K context inside that Mac’s default GPU limit. Estimates, not measurements.
| Build | Weights | RAM at 4K | at 32K | at 64K | Smallest Mac |
|---|---|---|---|---|---|
| As catalogued | — | 6 GB | 6 GB | 7 GB | 16 GB |
| Q2_K | 1.4 GB | 3 GB | 4 GB | 4 GB | 16 GB |
| Q3_K_M | 2.0 GB | 4 GB | 4 GB | 5 GB | 16 GB |
| Q4_K_M | 2.3 GB | 4 GB | 5 GB | 5 GB | 16 GB |
| Q5_K_M | 2.8 GB | 5 GB | 5 GB | 6 GB | 16 GB |
| Q6_K | 3.3 GB | 5 GB | 6 GB | 6 GB | 16 GB |
| Q8_0 | 4.2 GB | 6 GB | 7 GB | 7 GB | 16 GB |
| F16 | 8.0 GB | 10 GB | 10 GB | 11 GB | 16 GB |
Every published (filled square) and community (hollow circle) run the index holds for Nemotron 3.5 Bolt E4B: 0 published, 7 community, each linked to its source. LLMCheck runs no tests of its own. One row is one measurement session; on each chip the table above shows the lower median of the eligible runs (plain decode at 5 bits or fewer), and a binned or “not an anchor” run is shown but never used as a reference.
| Chip | Runtime | Quant | Decode | Context | Date | Source |
|---|---|---|---|---|---|---|
| M3 Max | MLX | 4-bit | community72.98 tok/s | short chat prompt, 256 generated, greedy | 2026-10-23 | huggingface.co binned |
| M3 Max | MLX | 6-bit | community55.22 tok/s | short chat prompt, 256 generated, greedy | 2026-10-23 | huggingface.co binned |
| M3 Max | MLX | 8-bit | community45.26 tok/s | short chat prompt, 256 generated, greedy | 2026-10-23 | huggingface.co binned |
| M3 Max | MLX | BF16 | community26.53 tok/s | short chat prompt, 256 generated, greedy | 2026-10-23 | huggingface.co binned |
| M5 Pro | MLX | oQ8e | community46.4 tok/s | 1k | 2026-10-23 | omlx.ai |
| M4 | MLX | OptiQ mixed 4/8-bit | community26.6 tok/s | code prompt, batch 1, greedy, thinking off, best of 3 | 2026-10-24 | huggingface.co |
| M4 | llama.cpp | Q4_K_M | community26.6 tok/s | short Japanese prompt, ~400 characters generated, -c 8192 | 2026-10-24 | note.com not an anchor |
The quickest start is Ollama. Pull the model by its tag (checked on ollama.com, 2026-11-14; the file is about 7.9 GB):
Prefer a window to a terminal? LM Studio downloads the same model with one click. · model card →
On the Mac mini M6 (16 GB), the smallest current Mac that runs it inside the default GPU limit, the index figure is ~42 tok/s (estimated) at Q4_K_M; keep the context at or under 64K there, or use a q8 KV cache. The nemotron-3.5-bolt:e4b file fits there.
Step-by-step: the Ollama install guide.
Mac mini M6 (16 GB) — smallest current Mac that runs it and cheapest at 15+ tok/s, $899 new (Apple list price, checked 2026-11-14), ~42 tok/s (estimated).
As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence the verdicts.
Yes. Nemotron 3.5 Bolt E4B needs about 6 GB of unified memory at a short (4K-token) context, so it runs on 53 of the 54 Mac configurations in the LLMCheck index. On 1 more it fits the RAM, but not even a safely raised GPU limit. The smallest Mac Apple sells new that runs it is the Mac mini M6 (16 GB).
About 6 GB at a 4K context: the weights plus about 2 GB for the runtime. Context adds KV cache: about 7 GB in total at 32K and 8 GB at 64K with the default 16-bit cache (about 7 GB at 64K with a q8 cache). These are estimates from the model’s config.json, not measurements.
On an M5 Max, the chip the leaderboard quotes it on, the index figure is ~161 tok/s (estimated) at Q4_K_M. The best measured figure is 46.4 tok/s on an M5 Pro 64 GB (community run, oQ8e). Every Mac’s figure is in the table on this page with its provenance: published, community run or estimated.
Among Macs Apple sells new, the Mac mini M6 (16 GB), $899 new (Apple list price, checked 2026-11-14), is the cheapest that reaches 15 tok/s with Nemotron 3.5 Bolt E4B, at ~42 tok/s (estimated).
On MacBook Air M1 8 GB (used) even a safely raised limit is short: expect heavy swapping, a CPU split or an out-of-memory error. From 16 GB up, every Mac Apple sells new holds it inside the default limit.
Q4_K_M (about 2 GB of weights) is the common 4-bit GGUF choice. Q8_0 (about 4 GB) is closer to full quality and needs a 16 GB Mac at the default GPU limit. Q3_K_M and Q2_K shrink it to about 2 and 1 GB at a quality cost. The index’s reference figure (161 tok/s on an M5 Max) is for its Q4_K_M build.
Its own window is 64K tokens. At macOS’s default GPU limit, the Mac mini M6 (16 GB) holds about 64K (the full window).
LLMCheck. (2026). Can I run Nemotron 3.5 Bolt E4B on a Mac? llmcheck.net. https://llmcheck.net/can-i-run/nemotron-35-bolt-e4b/ (updated ). CC BY 4.0.