furiosa-ai/Qwen3-VL-2B-Instruct

VISIONConcurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 8, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The furiosa-ai/Qwen3-VL-2B-Instruct is a 2-billion-parameter dense vision-language model from the Qwen3-VL series, developed by Qwen. It integrates a vision encoder with a dense transformer decoder, utilizing Interleaved-MRoPE and DeepStack for multimodal processing. This instruction-tuned model excels at visual understanding tasks including OCR, document analysis, spatial reasoning, and video comprehension, and natively supports tool calling. It is optimized for deployment on FuriosaAI RNGD hardware using Furiosa-LLM.

Loading preview...

Model Overview

The furiosa-ai/Qwen3-VL-2B-Instruct is a 2-billion-parameter dense vision-language model from the Qwen3-VL series, based on the upstream model by Qwen. It is specifically packaged with a Furiosa Executable Bundle (FXB) for efficient execution on FuriosaAI RNGD hardware using the Furiosa-LLM framework. This model combines a vision encoder with a dense transformer decoder, employing advanced techniques like Interleaved-MRoPE positional embeddings and DeepStack multi-level feature fusion to process both images and videos alongside text.

Key Capabilities

  • Vision-Language Understanding: Processes OpenAI-style multimodal chat messages, accepting image_url content parts alongside text for comprehensive visual understanding.
  • Multimodal Task Support: Capable of handling a wide range of visual tasks, including OCR, document and chart analysis, spatial reasoning, and video comprehension.
  • Native Tool Calling: Supports tool (function) calling via the hermes tool-call parser, enabling integration with external functionalities.
  • Optimized for FuriosaAI: Designed for deployment on FuriosaAI RNGD, running with a tensor-parallel size of 8 PEs per card.

Good For

  • Applications requiring robust multimodal AI capabilities on specialized hardware.
  • Developing solutions that need to interpret and respond to visual inputs (images, videos, documents).
  • Use cases involving tool-augmented language models for enhanced functionality and interaction.
  • Developers seeking an instruction-tuned vision-language model for deployment on FuriosaAI platforms.

Finetunes
1 models