Outmake Models
Models

Nemotron 3.5 Bolt E4B

NVIDIA · October 2026

Quick and steady with tools, NVIDIA's take on Gemma 4 E4B.

Rating, estimated
3.8
Downloads
792K
Parameters
8B
Context
64K
Download
3.5 to 15 GB
Commercial use
Yes

Runs great on M4 · 16 GB

Outmake picks the UD-Q4_K_XL size: a 5.1 GB download that uses 6.6 GB of memory, and writes about 11 words (15 tokens) a second.

Good at

  • Everyday
  • Fast

Licence

OpenMDW 1.1. Commercial use: yes. The maker's page has the full terms.

Requirements

Nemotron 3.5 Bolt E4B needs an Apple silicon Mac with 16 GB of memory or more. The smallest that runs it well is a MacBook Air, MacBook Pro, Mac mini or iMac with an M1 and 16 GB, on the UD-Q4_K_XL size: a 5.1 GB download that uses 6.6 GB of memory. There it writes about 6.4 words (8.6 tokens) a second. At 3 bits it fits 8 GB, with some loss in quality. 59 of 59 Mac configurations run it well.

Compare with

The nearest models in size, judged for the same Mac.

Sizes

SizeDownloadMemory at 8KMemory at 32KOpen
UD-Q3_K_XLSmaller4.2 GB5.3 GB5.6 GBOpen
UD-Q4_K_XLBalanced5.1 GB6.3 GB6.6 GBOpen
UD-Q6_K_XLBest7.3 GB8.7 GB9.0 GBOpen
All 9 sizes
SizeDownloadMemory at 8KMemory at 32KOpen
UD-IQ2_M3.5 GB4.6 GBestimated4.9 GBOpen
UD-IQ3_XXS3.7 GB4.8 GBestimated5.1 GBOpen
UD-Q2_K_XL3.7 GB4.8 GBestimated5.1 GBOpen
UD-Q5_K_XL5.8 GB7.1 GBestimated7.4 GBOpen
UD-Q8_K_XL8.2 GB9.7 GBestimated10.0 GBOpen
BF1615 GB17 GBestimated17 GBOpen

Memory is what llama.cpp's own calculator reported for each size when the catalog was built, with the app's settings. Smaller, Balanced and Best are the three sizes the store names.

Run It Anywhere

Nemotron 3.5 Bolt E4B is an open model, so any app that runs GGUF files can run it. These take the UD-Q4_K_XL size from unsloth straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF:UD-Q4_K_XL
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF:UD-Q4_K_XL
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF:UD-Q4_K_XL
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBUD-Q3_K_XL4.2 GB5.3 GB
8 GBUD-Q5_K_XL5.8 GB7.1 GB
10 GBUD-Q8_K_XL8.2 GB9.7 GB
11 GBUD-Q8_K_XL8.2 GB9.7 GB
12 GBUD-Q8_K_XL8.2 GB9.7 GB
16 GBUD-Q8_K_XL8.2 GB9.7 GB
20 GBBF1615 GB17 GB
24 GBBF1615 GB17 GB
32 GBBF1615 GB17 GB
48 GBBF1615 GB17 GB

The largest size, of those that keep three bits a weight or more, whose memory at an 8K chat is under 92% of the card's; the rest is the desktop's. Estimated from the files, the same way as for a Mac; speed on a PC is not estimated yet. A size that doesn't fit the card still runs, slower, with part of it in ordinary memory.

On every Mac
MacSizeFitMemoryWritesReads
M6 · 16 GBUD-Q4_K_XLRuns great6.6 GB16786
M6 · 24 GBUD-Q6_K_XLRuns great9.0 GB12786
M6 · 32 GBUD-Q6_K_XLRuns great9.0 GB12786
M5 · 16 GBUD-Q4_K_XLRuns great6.6 GB15290
M5 · 24 GBUD-Q6_K_XLRuns great9.0 GB11290
M5 · 32 GBUD-Q6_K_XLRuns great9.0 GB11290
M5 Pro · 24 GBUD-Q6_K_XLRuns great9.0 GB21580
M5 Pro · 48 GBUD-Q6_K_XLRuns great9.0 GB21580
M5 Pro · 64 GBUD-Q6_K_XLRuns great9.0 GB21726
M5 Max · 36 GBUD-Q6_K_XLRuns great9.0 GB301161
M5 Max · 48 GBUD-Q6_K_XLRuns great9.0 GB401451
M5 Max · 64 GBUD-Q6_K_XLRuns great9.0 GB401451
M5 Max · 128 GBUD-Q6_K_XLRuns great9.0 GB401451
M5 Ultra · 96 GBUD-Q6_K_XLRuns great9.0 GB722322
M5 Ultra · 256 GBUD-Q6_K_XLRuns great9.0 GB722902
M5 Ultra · 512 GBUD-Q6_K_XLRuns great9.0 GB722902
M4 · 16 GBUD-Q4_K_XLRuns great6.6 GB1198
M4 · 24 GBUD-Q6_K_XLRuns great9.0 GB8.398
M4 · 32 GBUD-Q6_K_XLRuns great9.0 GB8.398
M4 Pro · 24 GBUD-Q6_K_XLRuns great9.0 GB17195
M4 Pro · 48 GBUD-Q6_K_XLRuns great9.0 GB17195
M4 Pro · 64 GBUD-Q6_K_XLRuns great9.0 GB17195
M4 Max · 36 GBUD-Q6_K_XLRuns great9.0 GB24390
M4 Max · 48 GBUD-Q6_K_XLRuns great9.0 GB29488
M4 Max · 64 GBUD-Q6_K_XLRuns great9.0 GB29488
M4 Max · 96 GBUD-Q6_K_XLRuns great9.0 GB24390
M4 Max · 128 GBUD-Q6_K_XLRuns great9.0 GB29488
M3 · 8 GBUD-Q3_K_XLRuns well5.3 GB1181
M3 · 16 GBUD-Q4_K_XLRuns great6.6 GB9.581
M3 · 24 GBUD-Q6_K_XLRuns great9.0 GB781
M3 Pro · 18 GBUD-Q4_K_XLRuns great6.6 GB14141
M3 Pro · 36 GBUD-Q6_K_XLRuns great9.0 GB10141
M3 Max · 36 GBUD-Q6_K_XLRuns great9.0 GB19303
M3 Max · 48 GBUD-Q6_K_XLRuns great9.0 GB24404
M3 Max · 64 GBUD-Q6_K_XLRuns great9.0 GB24404
M3 Max · 96 GBUD-Q6_K_XLRuns great9.0 GB19303
M3 Max · 128 GBUD-Q6_K_XLRuns great9.0 GB24404
M3 Ultra · 96 GBUD-Q6_K_XLRuns great9.0 GB40605
M3 Ultra · 256 GBUD-Q6_K_XLRuns great9.0 GB40605
M3 Ultra · 512 GBUD-Q6_K_XLRuns great9.0 GB40807
M2 · 8 GBUD-Q3_K_XLRuns well5.3 GB1179
M2 · 16 GBUD-Q4_K_XLRuns great6.6 GB9.579
M2 · 24 GBUD-Q6_K_XLRuns great9.0 GB779
M2 Pro · 16 GBUD-Q4_K_XLRuns great6.6 GB18158
M2 Pro · 32 GBUD-Q6_K_XLRuns great9.0 GB13158
M2 Max · 32 GBUD-Q6_K_XLRuns great9.0 GB24297
M2 Max · 64 GBUD-Q6_K_XLRuns great9.0 GB24297
M2 Max · 96 GBUD-Q6_K_XLRuns great9.0 GB24376
M2 Ultra · 64 GBUD-Q6_K_XLRuns great9.0 GB40594
M2 Ultra · 128 GBUD-Q6_K_XLRuns great9.0 GB40594
M2 Ultra · 192 GBUD-Q6_K_XLRuns great9.0 GB40594
M1 · 8 GBUD-Q3_K_XLRuns well5.3 GB7.464
M1 · 16 GBUD-Q4_K_XLRuns great6.6 GB6.464
M1 Pro · 16 GBUD-Q4_K_XLRuns great6.6 GB16128
M1 Pro · 32 GBUD-Q6_K_XLRuns great9.0 GB12128
M1 Max · 32 GBUD-Q6_K_XLRuns great9.0 GB21220
M1 Max · 64 GBUD-Q6_K_XLRuns great9.0 GB21220
M1 Ultra · 64 GBUD-Q6_K_XLRuns great9.0 GB32441
M1 Ultra · 128 GBUD-Q6_K_XLRuns great9.0 GB32441

Writes and reads are words a second (a token is about three quarters of a word), 4K tokens into a chat, estimated unless marked measured. The size is the one the store offers that Mac. Constants fitted October 11, 2026; how, and how wrong.

Badge for the model card

Shows on Nemotron 3.5 Bolt E4B's Hugging Face card which Macs run it well, linked to this page. It updates every morning.

Runs on Mac badge

Sizes and files are read from unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF on Hugging Face and checked every morning. LM Arena has not rated this model, so the rating is Outmake's estimate from its size and release date; how the estimate is made. Catalog of November 15, 2026.

Your Computer

Every model is judged for this Mac. Both are under About This Mac, in the Apple menu.

Every Mac · Every graphics card

Open in OutmakeHugging Face

New to Outmake? Download the free beta