nemotron-3.5-bolt:e4b

13K 7 hours ago

Nemotron 3.5 Bolt is NVIDIA’s compact vision-language model for local coding, reasoning, and tool use.

vision tools thinking audio e4b 
ollama run nemotron-3.5-bolt:e4b

Details

7 hours ago

fe4af3f5907b · 7.2GB

json
{ "bos_token_id": 2, "do_sample": true, "eos_token_id": [ 1, 106, 50 ], "pad_token_id": 0, "temperat
json
{ "producer": { "name": "modelopt", "version": "0.48.0rc1.dev100+gf1d68fd4d" }, "quantization": { "q
json
{ "audio_ms_per_token": 40, "audio_seq_length": 750, "feature_extractor": { "dither": 0.0, "feature_
json
{ "version": "1.0", "truncation": null, "padding": null, "added_tokens": [ { "id": 0, "content": "<p
json
{ "audio_token": "<|audio|>", "backend": "tokenizers", "boa_token": "<|audio>", "boi_token": "<|imag
json
{ "architectures": [ "DFlashDraftModel" ], "model_type": "qwen3", "block_size": 8, "dflash_config":
json
{ "producer": { "name": "modelopt", "version": "0.47.0.dev69+g7b56dd6e3.d20260818" }, "quantization"
license
OpenMDW License Agreement, version 1.1 (OpenMDW-1.1) By exercising rights granted to you under this
params
{ "temperature": 1, "top_k": 64, "top_p": 0.95 }
json
{ "architectures": [ "Gemma4ForConditionalGeneration" ], "audio_config": { "_name_or_path": "", "arc
template
{{ .Prompt }}
2189 tensors

Readme

Nemotron 3.5 Bolt

Nemotron 3.5 Bolt is NVIDIA’s compact vision-language model for local coding, reasoning, and tool use.

Bolt E4B

Bolt E4B is an 8-billion-parameter model, with 4 billion effective, post-trained by NVIDIA from Gemma 4 E4B. It is designed for fast local conversations, instruction following, and multi-turn tool use. Its companion DFlash draft accelerates generation through speculative decoding. NVIDIA BF16 model card · NVIDIA Q4_K_M model card

ollama run nemotron-3.5-bolt:e4b

E4B is supported on both Windows/Linux and Apple Silicon. NVIDIA DFlash model card

Bolt 74B-A9B

It has 74 billion total parameters, with 9 billion active per token, and combines a Mamba-2/Transformer mixture-of-experts backbone with a C-RADIO vision encoder. NVIDIA model card

NVIDIA developed Bolt from Nemotron 3.5 Super through model compression and further training. Its published GGUF uses quantization-aware distillation for IQ2_XXS expert weights, alongside NVFP4 and Q8_0 for other layers.

  • Vision: Work with images, screenshots, charts, and documents.

  • Tools: Call functions and use their results in multi-turn tasks.

  • Thinking: Enable or disable reasoning before the answer.

  • Context: Supports up to 256K tokens. Usable context depends on available memory; images, conversation history, tool results, and output share that budget.

    ollama run nemotron-3.5-bolt

Compatibility

Apple Silicon / Metal is not currently supported for 74B-A9B. Memory requirements depend on context size, visual inputs, and speculative decoding.

License

Released under the OpenMDW License Agreement, version 1.1. See the NVIDIA model cards above for training details and evaluations.