What are the hardware requirements for Hy3?

Hy3 requires approximately 258.6 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 Hy3 and how large is it?

Hy3 is developed by tencent, a model with 298.8b 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 Hy3?

Hy3 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Hy3 by tencent — VRAM 258.6GB, 256K context

Name: Hy3
Author: tencent

Hy3 is an open-source language model by tencent. Features: 298.8b LLM, VRAM: 258.6GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.44.

Conversational Endpoints compatible Eval-results Hunyuan Hy3 Hy v3 Moe Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/tencent/Hy3

Hy3 Benchmarks

LLME Score: 0.44274

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

🌟 Advertise your project 🚀

Hy3 Parameters and Internals

LLM Name	Hy3
Repository 🤗	https://huggingface.co/tencent/Hy3
Model Size	298.8b
Required VRAM	258.6 GB
Updated	2026-07-08
Maintainer	tencent
Model Type	hy_v3
Model Files	7.2 GB: 1-of-99 7.2 GB: 2-of-99 7.2 GB: 3-of-99 7.2 GB: 4-of-99 7.2 GB: 5-of-99 1.1 GB: 6-of-99 7.2 GB: 7-of-99 7.2 GB: 8-of-99 7.2 GB: 9-of-99 7.2 GB: 10-of-99 7.2 GB: 11-of-99 1.1 GB: 12-of-99 7.2 GB: 13-of-99 7.2 GB: 14-of-99 7.2 GB: 15-of-99 7.2 GB: 16-of-99 7.2 GB: 17-of-99 1.1 GB: 18-of-99 7.2 GB: 19-of-99 7.2 GB: 20-of-99 7.2 GB: 21-of-99 7.2 GB: 22-of-99 7.2 GB: 23-of-99 1.1 GB: 24-of-99 7.2 GB: 25-of-99 7.2 GB: 26-of-99 7.2 GB: 27-of-99 7.2 GB: 28-of-99 7.2 GB: 29-of-99 1.1 GB: 30-of-99 7.2 GB: 31-of-99 7.2 GB: 32-of-99 7.2 GB: 33-of-99 7.2 GB: 34-of-99 7.2 GB: 35-of-99 1.1 GB: 36-of-99 7.2 GB: 37-of-99 7.2 GB: 38-of-99 7.2 GB: 39-of-99 7.2 GB: 40-of-99 7.2 GB: 41-of-99
Model Architecture	HYV3ForCausalLM
License	apache-2.0
Context Length	262144
Model Max Length	262144
Transformers Version	5.6.0
Vocabulary Size	120832

Quantized Models of the Hy3

Model	Likes	Downloads	VRAM
Hy3 GPTQ RTN 4bit Tp8	2	46	161 GB
Hy3 MLX Q9	2	27	330 GB

Best Alternatives to Hy3

Best Alternatives	Context / RAM	Downloads	Likes
Hy3 Preview	256K / 221.4 GB	74431	297
Hy3 FP8	256K / 129.6 GB	11017	54

Note: green Score (e.g. "73.2") means that the model is better than tencent/Hy3.

Rank the Hy3 Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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

What open-source LLMs or SLMs are you in search of? 55235 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Hy3 by tencent