# NVIDIA's Nemotron-3.5 Bolt targets 24GB GPUs

> A multimodal mixture-of-experts model with 74B total parameters but just 9B active, tuned to fit consumer-class hardware.

Published by The Open Weights on Oct 7, 2026. Canonical: https://theopenweights.com/news/nemotron-3-5-bolt-74b-a9b-gsz4

## Key facts

- Company: NVIDIA
- Model: Nemotron-3.5 Bolt 74B-A9B
- Version: Nemotron-3.5 Bolt
- Category: Vision-Language
- Modalities: Text / LLM, Vision-Language
- License: Other (Open weights, commercial use allowed)
- Parameters: 74B (mixture of experts)
- Active parameters: 9B
- Context window: 262K tokens
- Architecture: NemotronH_Omni_Reasoning_V3
- Precision: INT8
- Significance: notable
- Published: 2026-10-07
- Last verified: 2026-10-20
- Hugging Face: https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB
- Canonical URL: https://theopenweights.com/news/nemotron-3-5-bolt-74b-a9b-gsz4

NVIDIA has published Nemotron-3.5 Bolt, a multimodal mixture-of-experts model that pairs a large 74-billion-parameter capacity with a lean nine-billion active footprint. The release is available now on [Hugging Face](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB).

The naming is unusually explicit about its intended home: the "24GB" suffix signals that NVIDIA has sized this model to run on a single 24GB GPU, the class of memory found in cards like the RTX 4090 and professional workstation parts. That framing matters because large vision-language models have typically demanded multi-GPU setups or datacenter accelerators.

## Why the MoE math matters

Mixture-of-experts designs route each token through only a fraction of the network, so the compute and memory cost at inference tracks the active parameters rather than the full total. Here that means:

- 74B total parameters for overall capacity
- 9B active parameters per forward pass
- A target of 24GB of GPU memory

That combination is the central pitch — expert-level breadth without the full hardware bill. The model is listed as a vision-language model, extending Nemotron beyond text into multimodal understanding.

NVIDIA has not published a context length in this record, and the model carries a custom "other" license, so teams will want to review the terms before building on it. Still, a multimodal MoE engineered for consumer-class memory lowers the barrier for developers who want capable vision-language inference on hardware they already own.

## Get the model

- [Hugging Face](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB)

## Sources

- [nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB) — Hugging Face, Oct 7, 2026

---
Source: The Open Weights (https://theopenweights.com/). Aggregated and written by Claude, curated by humans. Cite as: "NVIDIA's Nemotron-3.5 Bolt targets 24GB GPUs", The Open Weights, Oct 7, 2026, https://theopenweights.com/news/nemotron-3-5-bolt-74b-a9b-gsz4