The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestTencentHy4
TencentText / LLM

Tencent releases Hunyuan Hy4, an MoE vision-language model

The instruction-tuned, mixture-of-experts model handles both images and text under an open weights listing on Hugging Face.

Oct 29, 2026
NotableOther
Hunyuan Hy4

Tencent has published Hunyuan Hy4, a multimodal vision-language model, to Hugging Face. The release extends the company's Hunyuan line into territory that pairs image understanding with text generation in a single system.

Hy4 is described as an instruction-tuned model built on a mixture-of-experts (MoE) architecture. MoE designs activate only a subset of their parameters for any given input, a technique that lets developers scale total capacity while keeping the compute cost of each inference in check. That approach has become a common thread across recent frontier-scale open releases.

Why it matters

Vision-language models are the connective tissue for a growing set of applications — document parsing, UI automation, visual question answering, and multimodal assistants among them. A capable open-weights VLM from a company of Tencent's scale gives developers another foundation to build on without committing to a closed API.

A few practical notes from the listing:

  • It is positioned as a vision-language model with text generation capability.
  • It uses a mixture-of-experts architecture and is instruction-tuned.
  • It ships under a custom license, so teams should review the terms before commercial use.

Tencent has not yet detailed parameter counts, context length, or benchmark results in the record available at launch. Prospective users will want to watch the model card for documentation, evaluation numbers, and licensing specifics as they are published.

Sources

  • tencent/Hy4

    Hugging Face

    Visit
OlderKRAFTON Releases Raon-1-nano, a Sub-1B Open ModelKRAFTON · Text / LLM · 3 weeks agoNewerIBM's Granite Switch Treats Models Like SoftwareIBM · Any-to-Any · 3 weeks ago

Get the model

Hugging Face

Specs

Context window1.0M tokens
Size1.5 TB
PrecisionBF16
ArchitectureHYV4VLForConditionalGeneration
LicenseOTHER
Downloads11.1K
Likes564

Can you run it?

Datacenter only — about 466.1 GB even at 4-bit.

  • BF16 (as published)

    Multi-node cluster

    1.6 TB
  • 8-bit

    Multi-node cluster

    822.2 GB
  • 4-bit

    8×H100 80 GB node

    466.1 GB
Your machine
GGUF builds
  • Mixture-of-experts: every expert must be loaded, so memory follows total parameters, not the active slice.
  • KV cache sized for 8,192 tokens of context; longer prompts need proportionally more.

Based on Hugging Face weights metadata. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

Text / LLMVision-Language
2 versions — view changelog

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More from Tencent

All Tencent releases →
Hunyuan Hy4-flash
Tencent/Vision-LanguageMulti-GPU server

Tencent releases Hunyuan Hy4-flash multimodal MoE

The lightweight, mixture-of-experts model adds vision and text instruction-following to Tencent's open Hunyuan lineup.

Nov 17, 2026
Hy-Nomas
Tencent/Text → Video

Tencent Releases Hy-Nomas, an Interactive Video World Model

The open-weights model pairs text- and image-to-video generation with action-aware causal attention and explicit camera control.

Oct 21, 2026
Hy-Memory 30B-A3B
Tencent/Text / LLMConsumer GPU

Tencent releases Hy-Memory 30B-A3B under Apache 2.0

A memory-focused mixture-of-experts model activates just 3B of its 30B parameters per token, and ships with a permissive license.

Oct 15, 2026

More in Text / LLM

All Text / LLM →
Aion-1.0-Instruct-2.3B
Microsoft/Text / LLMRuns anywhere

Microsoft's Aion 1.0 Instruct is a compact 2.3B chat model

A small, multilingual instruction-tuned model aimed at chat and summarization lands on Hugging Face.

Nov 18, 2026
SenseNova-Loop-9B
SenseTime/Vision-LanguageRuns on a laptop

SenseTime Releases Looped 9B Multimodal Model

The company's new SenseNova-Loop-9B pairs vision-language capabilities with a recurrent looped-transformer design.

Nov 17, 2026
Sage S2.1 Flash
Internlm/ReasoningMulti-GPU server

InternLM's Sage S2.1 Flash targets scientific reasoning

The 320B mixture-of-experts model is built to reason over scientific evidence and act inside interactive environments.

Nov 15, 2026