Qwen3.6-35B-A3B-NVFP4 is a foundation model hosted on Hugging Face under the nvidia organization. The model processes multimodal inputs that include text, images, and video through a specialized chat template. Its template defines distinct handling for each content type, inserting vision-specific tokens such as vision_start, image_pad, video_pad, and vision_end while counting vision elements and raising exceptions for unsupported cases like videos in system messages.
The template also supports tool calling by generating a system prompt that lists available functions when tools are supplied. It iterates over conversation messages, applies conditional formatting based on content type, and enforces rules such as requiring at least one message. The implementation appears as Jinja2-style macros that output structured strings compatible with the model's expected input format.
This model is listed alongside standard Hugging Face infrastructure for models, datasets, spaces, and community resources. The associated organization page promotes open source and open science initiatives in artificial intelligence.
In the Foundation models & chat space, Qwen3.6 35B A3B takes a focused approach. It focuses on enabling advanced text and multimodal AI applications with a large, open-source model. It is built as an open-source project for AI developers and researchers needing large-scale language models. Qwen3.6 35B A3B is open source under the Apache-2.0 license. Qwen3.6 35B A3B is available on the web, the command line, and API, and it can be self-hosted.
It is developed by NVIDIA, and the product first shipped in 2024. Development happens publicly on GitHub with 3.3k stars and 357 commits in the last 90 days. The category is crowded — PulseGate's index counts 25 comparable apps. Key capabilities include large language model, multimodal support, and text generation.
Latest indexed changes and source events
nvidia/Qwen3.6-35B-A3B-NVFP4 verified by the PulseGate indexer
Other apps tracked under the same category.
