DownloadNVIDIA Nemotron 3.5 Bolt 74B-A9B 24GB is an open-weight vision-language model optimized for running autonomous AI agents locally within a 24GB memory footprint. Compressed from Nemotron 3.5 Super VL using NVIDIA's Star Elastic method, it contains 74B total parameters with 9B active per token. Its hybrid Mamba-2–Transformer LatentMoE architecture integrates a C-RADIO v4-H vision encoder for understanding text, images, and video. The model supports a 256K-token context window and uses native IQ2 quantization with quantization-aware distillation to preserve accuracy. It targets everyday agentic workflows involving coding, web browsing, terminal operations, and tool calling. Configurable thinking modes balance direct responses and deeper reasoning, while LiLiCorr speculative decoding improves generation speed. Designed for NVIDIA GPU systems, it supports vLLM deployment and commercial or noncommercial use under the OpenMDW 1.1 license.
Features
- 74B total parameters with 9B activated per token
- Hybrid Mamba-2–Transformer LatentMoE architecture
- C-RADIO v4-H vision encoder for image and video understanding
- 256K-token context window for extended agent workflows
- Native IQ2 quantization for a 24GB memory footprint
- Star Elastic compression with quantization-aware distillation
- Configurable thinking modes and LiLiCorr speculative decoding
- Optimized for local coding, web browsing, terminal use, and tool calling
Project Samples
