Nemotron 3.5 Bolt E4B 4GB: An Open and Efficient Small Language Model for Fast Conversational Agents
We introduce Nemotron 3.5 Bolt E4B 4GB, an open and efficient small language model further post-trained from Gemma-4-E4B-it and optimized for speed and intelligence per GB of memory used for fast conversational agents. The model carries forward Gemma's strong conversational capabilities and further improves multi-turn tool use, instruction following, and input grounding for fast, local voice/device assistants, interactive personas, and embedded/Edge interfaces. The model combines natural conversation with multi-turn agentic capabilities to power voice/device assistants, interactive characters and robotic/IoT interfaces running on GeForce RTX, Jetson, and RTX/DGX Spark. Experiments on 21 benchmarks show that Nemotron 3.5 Bolt E4B 4GB achieves the highest average accuracy while generating the fewest output tokens in both reasoning and non-reasoning modes, compared with Gemma-4-E4B-it and Qwen3.5-4B. The model is released with a DFlash drafter. With DFlash, Bolt E4B achieves $2.74\times$ the per-user decode speed of inference without speculative decoding at concurrency~1, and $1.21\times$ the generation throughput of Gemma-4-E4B-it with MTP at batch size 1.
Authors
- Ashwath Aithal
- L. Dykes
- Aishwarya Chhabra
- Oluwatobi Olabiyi
- Michael Hasper
- Jian Leong
- Besmira Nushi (ORCID: https://orcid.org/0000-0002-7554-8586)
- Giedrius Burachas
- Leili Tavabi (ORCID: https://orcid.org/0009-0003-5270-1984)
- Felix Friedmann
- Aditya Malte
- Grzegorz Chlebus (ORCID: https://orcid.org/0000-0001-6994-696X)
- Shaona Ghosh (ORCID: https://orcid.org/0000-0003-4658-5174)
- Bhavesh Pawar (ORCID: https://orcid.org/0009-0001-8332-3854)
- Manoj Kilaru
- Annamalai Chockalingam
- Andrii Skliar
- Ambrish Dantrey
- Ruvaid Virk (ORCID: https://orcid.org/0000-0002-6292-8264)
- Oleksii Kuchaiev
- Prasoon Varshney (ORCID: https://orcid.org/0009-0008-8449-7400)
- Eileen Long (ORCID: https://orcid.org/0000-0002-0743-4464)
- Jiaqi Zeng (ORCID: https://orcid.org/0000-0002-6967-5394)
- Alex Hodson
- Krzysztof Pawelec
- Ameya Sunil Mahabaleshwarkar
- Saeid Motiian
- Bryan Catanzaro (ORCID: https://orcid.org/0000-0003-0034-7728)
- Haifeng Qian (ORCID: https://orcid.org/0000-0002-7189-6903)
- Erick Galinkin (ORCID: https://orcid.org/0000-0003-1268-9258)
- Ruby Ye (ORCID: https://orcid.org/0000-0001-6803-5494)
- Nancy Agarwal (ORCID: https://orcid.org/0000-0003-4392-0520)
- Hayley Ross (ORCID: https://orcid.org/0000-0003-2288-3847)
- David Mosallanezhad
- Ryan Hill (ORCID: https://orcid.org/0009-0009-8277-0458)
- Pablo Ribalta
- J. Vialard
- NVIDIA
- Payal Motwani
- Seph Mard
- Kari Briski
- Dina Yared
- Thinh Le
- Guillermo Siman
- Amy Shen
- Jonathan Cohen
- Gorkem Batmaz
- Katherine Luna
- Ran Zilberstein
- Thanh Do
- Terry Kong
- Chris Alexiuk
- Negar Habibi
- Meredith Price
- Pinky Xu
- Chantal Hwang
- Richard Mazzarese
- Micah Schaffer
- Michael Ross Boone (ORCID: https://orcid.org/0009-0003-2086-6856)
- Michael Fukuyama
- Sudeep Sabnis
- Pritam Biswas
- Rameshwar Shivbhakta
- Joey Conway
- Rita Fernandes Neves
- Reka Kovacs
- Lily Lee
- Adi Margolin
- Puhui Meng
- Ehsan Hosseini Asl
- Rochelle Woods
- Shai Ben Nun
- Jose Vergara
- Maka Dong
- Linda Liu
- Chenhan Yu
- Khalil Ben Khaled
- Conner Currie
- Kevin Ning
- Sharad Yadav
- Irina Filatova
- Rishabh Garg
- Jian Zhang
- Suhas Hariharapura Sheshadri
- Kedar Pandurang Potdar
- Matthew Lee
- Ike Nnoli
- Huiying Li
- Anusha Pant
- Hao Guo
- Iris Liu
- Ashwin Kumar
- Hrishith Thadicherla
- Ji-Eun Lee
- Bryan Yu
- Cynthia Gao
- Maria Bagdasarova
- Liwei Ding
- Hui Wang
Publication Details
- Journal
- arXiv (Cornell University)
- Published
- 2026-11-16
- DOI
- https://doi.org/10.48550/arxiv.2611.20771
- Primary Topic
- Topic Modeling
- Type
- preprint