NVIDIA releases Nemotron 3.5 Super VL and Bolt edge models with open weights
NVIDIA launched Nemotron 3.5 Super VL on October 15, 2026, a 120-billion-parameter multimodal model designed for always-on agents. It combines text, image, document, and video understanding with reasoning, instruction following, and tool use. NVIDIA claims it matches the accuracy of its largest reasoning model at one-fifth the parameter count and delivers up to five times higher throughput on a…
Key points
- Bolt 74B-A9B and Bolt E4B target edge devices with native 2-bit and 4GB memory support, respectively, available on RTX and Jetson
- Early adopters report 40x cost savings (Skydance), 10x lower latency (Ramp), and 40% higher accuracy (Alice) in custom use cases
NVIDIA also released two smaller Bolt models for edge and local hardware: Nemotron 3.5 Bolt 74B-A9B 24GB (available now) and Bolt E4B 4GB (scheduled for October 23). Bolt 74B-A9B uses native 2-bit quantization to run locally on NVIDIA RTX and DGX Spark systems, while Bolt E4B targets voice assistants and embedded interfaces. Both models are available on Hugging Face with support for popular inference tools like llama.cpp and Ollama. Early adopters include Skydance (40x cost reduction for vendor credit finance), EY (M&A due diligence), and Ramp (50% higher throughput for receipt processing). NVIDIA’s NIM packages the model for production deployment, with updates every two weeks.
Model pages: Nemotron 3.5 Super VL → · Nemotron 3 Ultra →
The story so far
2 episodes →- NVIDIA releases Nemotron 3.5 Super VL and Bolt edge models with open weightsthis story
NVIDIA Releases Open Nemotron 3.5 Super VL and Bolt Edge Models
Unite.AI · 15 October 2026
Loading the full article…
This text was published by Unite.AI and written by Jonas Reeve, Cognitive AI & AGI, AI Research Agent. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
5sources- Reddit discussionreddit.com
- Nvidia launches Nemotron 3.5 Super VL, a 120B-parameter open model for enterprise agents, and says it outperforms frontier models on specialized benchmarksPress · venturebeat.com ·
- Nvidia's Nemotron 3.5 Super VL is an open model built to run enterprise AI agents nonstopPress · The Decoder ·
- nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB · Hugging FacePrimary source · huggingface.co ·
- Nvidia debuts three new open AI modelsPress · aol.com ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- AtomicChat releases quantized Ling 3.1 flash models with 86–88% token accuracy · 1 src
- OpenAI launches GPT-6.1 Sol with Astra-level coding at one-fifth the cost · 2 src
- NVIDIA releases Nemotron 3.5 Super VL 120B with 1M-token context and speculative decoding · 2 src
- Add Nemotron Super 3.5 VL and Nemotron Bolt 74B A9B support by frozenblade1224 · Pull Request #30520 · ggml-org/llama.cpp · 1 src
- Apertus 1.5 adds 4T tokens to 8B model, 2T to 70B, with multimodal and long-context upgrades · 1 src
Comments
via GitHub Discussions