Qwen3.5 122B A10B Alternatives

Qwen3.5-122B-A10B-NVFP4 is a quantized version of the Qwen3.5 multimodal model published by RedHatAI on Hugging Face. It supports vision and language capabilities with optimized performance for NVIDIA hardware using NVFP4 precision. The model is intended for developers integrating advanced multimodal AI into their applications or research.

Qwen3.6-27B-NVFP4 is an NVIDIA-optimized version of the Qwen3 27B model using NVFP4 quantization. It supports vision and video inputs in addition to text and includes a sophisticated tokenizer and chat template. The model is designed for efficient inference on NVIDIA GPUs while maintaining the strong performance of the original Qwen3 architecture.

Qwen3-30B-A3B-NVFP4 is an NVIDIA-optimized version of the Qwen3 model using NVFP4 quantization. It supports advanced tool calling and function calling through a detailed chat template. The model is designed for efficient inference on NVIDIA GPUs while preserving the capabilities of the original Qwen3 architecture.

Qwen3.5-122B-A10B-NVFP4 is a highly quantized version of the large Qwen 3.5 model (122B parameters). It supports vision inputs alongside text and includes advanced tool-calling features. The model is distributed on Hugging Face for use in high-performance inference environments.

Qwen3-32B-NVFP4 is an FP4 quantized version of the Qwen3-32B model created by NVIDIA. It includes support for tool calling and is optimized for high-performance inference on NVIDIA GPUs. The model is distributed on Hugging Face for developers seeking state-of-the-art performance with reduced memory and compute requirements.

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.

This NVIDIA-hosted quantized version of the Qwen 3.5 397B (with 17B active parameters) model supports vision and video inputs in addition to text. It includes advanced prompting templates for tool use and multimodal content. The model is designed for high-performance inference using NVIDIA-optimized stacks.

nvidia/Qwen3-8B-NVFP4 provides a quantized version of the Qwen3 8B model optimized for NVIDIA GPUs. It supports advanced features such as tool calling and is designed for high-performance local or server-side inference. The model is distributed via Hugging Face.

This repository contains an NVIDIA-optimized FP4 (NVFP4) quantized version of the Qwen3 14B model. It includes specialized chat templates and tool-calling support optimized for NVIDIA inference stacks. The quantization enables faster and more memory-efficient inference while preserving model capabilities.

This is a hybrid int4/fp8 quantized version of the Qwen3.5-122B-A10B model with full support for image and video inputs. It features sophisticated content rendering for multimodal messages and vision counting. The model balances performance and efficiency for large-scale multimodal applications.

This is a highly optimized, quantized version of the Qwen 3.6 35B model using NVFP4 precision for accelerated inference. It supports both text and vision inputs while maintaining strong performance. Distributed via Hugging Face, it is designed for developers needing fast multimodal inference on consumer or enterprise GPUs with Unsloth and Transformers compatibility.

This is a community-quantized version of the Qwen3.6-35B model using NVFP4 precision. It supports both text and vision inputs and is optimized for reduced memory usage and faster inference on compatible hardware. The model is intended for local deployment and experimentation.

This repository contains a w8a8 quantized version of the Qwen3.5-9B model from RedHatAI. It supports both text and vision inputs and is optimized for lower memory usage and faster inference. The model includes vision tower components and is compatible with the Transformers library.

Qwen3.5-9B-NVFP4 is a quantized (NVFP4) version of Alibaba's Qwen3.5-9B large language model. It is optimized for reduced memory usage and faster inference while maintaining strong performance. The model supports multimodal inputs and can be run locally using standard Hugging Face tools.

Qwen3-32B-FP8 is an FP8-quantized version of Alibaba's Qwen3 32B large language model. It supports advanced capabilities including tool calling, long-context understanding, and multilingual performance while significantly reducing VRAM usage compared to the original. The model is designed for local and self-hosted inference by developers building AI applications.

Qwen3.6-35B-A3B-NVFP4 is an open-source large language model variant designed for advanced reasoning and tool calling. It is suitable for developers and researchers seeking a high-capacity LLM for integration into AI applications or research workflows. The model supports both local and API-based inference and is distributed with open weights.

Qwen3.6-27B-FP8 is a community-quantized variant of Alibaba's Qwen3 model offered in FP8 precision for lower memory usage while retaining strong performance. It supports text and vision inputs and is compatible with the Hugging Face ecosystem for local or cloud deployment. It is intended for developers building efficient multimodal applications.

Qwen3.5-122B-A10B-GGUF is a quantized release of the Qwen3.5-122B-A10B foundation model hosted on Hugging Face by Unsloth. It is provided in GGUF format for local inference. The model supports processing of text, image, and video inputs through a chat template that handles multimodal content. Its template includes specific tokens such as vision_start, vision_end, image_pad, and video_pad, along with logic for counting vision elements and raising exceptions for unsupported cases like videos in system messages. The template also accommodates tool calling by formatting available functions when tools are supplied in the messages. It is delivered as a repository on the Hugging Face platform containing GGUF quantized files. The page includes a chat template implementation in a macro-based format that processes message lists, handles different content types, and generates appropriate system prompts for tool use. The model belongs to the class of foundation models. No pricing, licensing details, or target audience beyond the repository context are stated.

This is an AWQ-quantized 27B parameter version of Alibaba's Qwen3.5 language model. It supports text generation, tool use, and multimodal inputs while requiring significantly less memory than the original. The model is distributed via Hugging Face for use with popular inference frameworks.

Qwen3.6-27B-NVFP4 is a community-quantized version of the Qwen3.6 27-billion parameter model using NVFP4 precision. It includes patches for improved coding-agent and tool-calling behavior. The model is hosted on Hugging Face and designed for efficient local or server-based inference on compatible NVIDIA hardware.

Qwen3-30B-A3B-GPTQ-Int4 is a GPTQ 4-bit quantized variant of the Qwen3 30B-A3B model. It supports advanced capabilities such as tool calling and follows a detailed chat template for conversational use. The model is optimized for local or server-based inference using tools like llama.cpp or Hugging Face Text Generation Inference while maintaining strong performance.

This is an INT4 GPTQ-quantized version of the Qwen3.5-27B model from the Qwen team. It supports multimodal inputs including text, images, and video. The quantization enables more efficient inference while maintaining strong performance across various tasks.

A heavily quantized (4-bit AWQ) version of a Qwen 3.6 model with 35B+3B parameters and vision capabilities (VRAP). The 21.2GB model supports both text and image inputs. It is designed for local inference using GGUF-compatible tools or optimized runtimes.

A 4 billion parameter version of the Qwen3 large language model in FP8 precision. It supports instruction following, tool calling, and general text generation. The model is hosted on Hugging Face and can be used locally or via inference providers.

Qwen3-1.7B-FP8 is a quantized variant of Alibaba's Qwen3 series of large language models. It supports advanced features including tool calling and follows a chat template optimized for instruction following. The FP8 format enables faster inference with reduced memory requirements while maintaining strong performance.

Qwen3-30B-A3B-FP8 is an FP8 quantized version of Alibaba's Qwen3 series Mixture-of-Experts model. It offers strong performance on reasoning and tool-use tasks while significantly reducing memory requirements compared to full-precision versions. The model supports advanced features such as function calling and multi-step tool use and is distributed for use with Transformers and compatible inference engines.

This is a 4.75-bit PrismaQuant version of a Qwen3.6 35B-A3B model optimized for use with the vLLM inference engine. It supports multimodal inputs including vision and includes custom chat templates for tool use. The model is distributed on Hugging Face for efficient local or server-based deployment.

Qwen3.6-27B-NVFP4 is an FP4 quantized variant of the Qwen3.6-27B model. It supports both text and vision inputs including images and video through specialized tokens and templates. The model is designed for efficient local or hosted inference using the Transformers library and is available on the Hugging Face model hub.

This repository hosts an Intel-optimized, AutoRound-quantized version of the Qwen3.5-122B-A10B model in int4 precision. It enables running this large multimodal model on more accessible hardware while maintaining good performance. The model is suitable for advanced AI research and local inference applications.

This is a 4-bit AWQ quantized version of Alibaba's Qwen3.5-122B (A10B) model. Hosted on Hugging Face, it enables local inference of a powerful 122 billion parameter language model on consumer hardware. The model supports advanced features including tool calling and multimodal inputs.

Qwen3.6-27B-NVFP4 is an open-source large language model supporting both text and vision modalities. It can be deployed locally or in the cloud, making it suitable for AI researchers and developers who need flexible, high-capacity models for advanced NLP and computer vision tasks.

Qwen3.6-35B-A3B-AWQ-4bit is a quantized version of a large language model, designed for efficient inference on local hardware. It is suitable for developers and researchers who need high-performance language models with reduced resource requirements.

Qwen3.5-35B-A3B is an open-source large language model supporting both text and multimodal inputs. It is designed for advanced AI applications, including chatbots and multimodal assistants, and is suitable for developers and researchers in AI.