What are the hardware requirements for Qwen3xd?

Qwen3xd requires approximately 4 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Qwen3xd and how large is it?

Qwen3xd is developed by cyrah-ai, a model with 2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Qwen3xd?

Qwen3xd is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Qwen3xd by cyrah-ai — VRAM 4GB, 256K context

Name: Qwen3xd
Author: cyrah-ai

Qwen3xd is an open-source language model by cyrah-ai. Features: 2b LLM, VRAM: 4GB, Context: 256K, License: other, Merged, LLM Explorer Score: 0.2.

Merged Model Conversational En Endpoints compatible Es Experimental Qwen3 moe Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/cyrah-ai/qwen3xd

Qwen3xd Parameters and Internals

LLM Name	Qwen3xd
Repository 🤗	https://huggingface.co/cyrah-ai/qwen3xd
Merged Model	Yes
Model Size	2b
Required VRAM	4 GB
Updated	2026-07-09
Maintainer	cyrah-ai
Model Type	qwen3_moe
Model Files	4.0 GB
Supported Languages	en es
Model Architecture	Qwen3MoeForCausalLM
License	other
Context Length	262144
Model Max Length	262144
Transformers Version	4.57.0.dev0
Vocabulary Size	262170

Rank the Qwen3xd capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Qwen3xd

Best Alternatives	Context / RAM	Downloads	Likes
OpenAster1 128K Base	128K / 4.1 GB	9	0

Note: green Score (e.g. "73.2") means that the model is better than cyrah-ai/qwen3xd.