{"slug": "model-nemotron-3-5-bolt-e4b", "title": "Model: Nemotron 3.5 Bolt E4B", "summary": "NVIDIA released Nemotron 3.5 Bolt E4B, a small vision model built on Google's Gemma 4 E4B with the same 42-layer shape, about 4.7B working parameters plus a 2.8B per-layer embedding table, image input and a 64K context window. Only 4 of the 42 layers keep a context-growing cache, and the model runs in roughly 6.6 GB at 4K context and 7.6 GB at 64K in its 6.2 GB Q4_K_M file, while the Q8_0 8.0 GB file needs about 8.5 GB at 4K and a 12 GB card. The vision encoder is a separate 0.6–1.0 GB file published by unsloth, and NVIDIA's benchmark claims are its own with no per-level quality-loss data published.", "body_md": "NVIDIA's Nemotron 3.5 Bolt E4B is a small vision model built on Google's Gemma 4 E4B. It has the same 42-layer shape, about 4.7B parameters doing the work plus a 2.8B per-layer embedding table, with image input and a 64K window. Only 4 of its 42 layers keep a cache that grows with context, so long context is cheap. The one plain Q4_K_M file (LM Studio's, 6.2 GB) is much larger than the level name suggests, because the embedding table is not stored at 4 bits. It needs about 6.6 GB at 4K, comfortable on an 8 GB card, and 7.6 GB at 64K, tight on an 8 GB card. Q8_0 (an 8.0 GB file) needs about 8.5 GB at 4K and wants a 12 GB card. On a discrete card llama.cpp keeps the embedding table in system RAM, so real VRAM use is lower than shown. The vision encoder is a separate 0.6–1.0 GB file published by unsloth, and a small DFlash draft model for speculative decoding ships beside it. NVIDIA's benchmark claims are its own, and no per-level quality loss has been published.", "url": "https://wpnews.pro/news/model-nemotron-3-5-bolt-e4b", "canonical_source": "https://quantized.uk/quant-hub/nemotron-3.5-bolt-e4b/", "published_at": "2026-10-24 11:41:15.511044+00:00", "updated_at": "2026-10-24 11:41:17.791952+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "computer-vision", "ai-research"], "entities": ["NVIDIA", "Nemotron 3.5 Bolt E4B", "Google", "Gemma 4 E4B", "unsloth", "llama.cpp", "LM Studio"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/model-nemotron-3-5-bolt-e4b", "markdown": "https://wpnews.pro/news/model-nemotron-3-5-bolt-e4b.md", "text": "https://wpnews.pro/news/model-nemotron-3-5-bolt-e4b.txt", "jsonld": "https://wpnews.pro/news/model-nemotron-3-5-bolt-e4b.jsonld"}}