丽蟾CGC 
Qwen3.6-35B-A3B-Spark NIM Certified
Multimodal MoE model supporting text/image/video inputs with OpenAI-compatible API, powered by SGLang.
MiniMax-M2.5 Spark
MiniMax-M2.5 is a text generation model optimized for complex agentic tasks including software engineering, tool use, search, and office-work style workflows.
Gemma-4-26B-A4B-IT-Spark
Gemma-4-26B-A4B-IT-Spark NIM 容器打包了 Google gemma-4-26B-A4B-it 模型，以便通过 NVIDIA NGC 作为可下载的 NIM 进行部署。 该容器组件已准备好投入商业使用。
Qwen3.6-35B-A3B-Spark NIM Container Overview
Multimodal MoE model supporting text/image/video inputs with OpenAI-compatible API, powered by SGLang.
NIM Base Container Spark
Enterprise-grade base container for building custom LLM NIMs with SGLang v0.5.9, secure & high-performance
NIM-BaseContainer-Spark
NIM
Build an Enterprise RAG Pipeline Blueprint
Power fast, accurate semantic search across multimodal enterprise data with NVIDIA’s RAG Blueprint—built on NeMo Retriever and Nemotron models—to connect your agents to trusted, authoritative sources of knowledge.
Build a Video Search and Summarization (VSS) Agent
Ingest massive volumes of live or archived videos and extract insights for summarization and interactive Q&A
Isaac Sim
NVIDIA Isaac Sim™ is a robotics and AI simulation application framework built on NVIDIA Omniverse™. Isaac Sim has essential features for building virtual robotic worlds and experiments.
nvidia-nemotron-nano-9b-v2-dgx-spark
NVIDIA-Nemotron-Nano-9B-v2 is a language model that can follow instructions, complete requests, and generate creative text formats. The NVIDIA-Nemotron-Nano-9B-v2 Large Language Model (LLM) is an instruct fine-tuned version of the Nemotron-Nano.
qwen3-32b-dgx-spark
Qwen3-32B NIM for DGX Spark
Llama-3.1-8b-Instruct-DGX-Spark
Llama 3.1 8B is optimized for dialogue and conversational use cases, making capable of following user instructions to perform a wide variety of natural language processing tasks. This container is designed to run on NVIDIA DGX Spark.
Qwen3-30B-A3B-NVFP4-spark
Qwen3 30B A3B NVFP4 NIM for DGX Spark
- 1