Tencent

Hy4 preview

CTX1.02M

Preview

88Quality

Reasoning · Tools

Input$0.83/M

Output$2.50/M

A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.

Open weightsTencent Hunyuan license991/1044

Drag to turn

TencentTextUncommonPreview

Hy4 preview

A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.

hy4-preview

Type: Text
Quality: 88 · Top 12% on the LMArena WebDev
Context window: 1,024,000 tokens
Max output: 64,000 tokens
Input price: $0.83 / 1M tokens
Output price: $2.50 / 1M tokens
Cache read: $0.042 / 1M tokens
Cache write: —
Parameters: 780B (779,960,992,733)
Downloads: 22K in the last 30 days
Likes: 510
Access to weights: Open download
Released: Aug 28, 2026
Knowledge cutoff: —
Takes in: Text
Puts out: Text
Reasoning: Yes
Tool calling: Yes
Structured output: No
File attachments: No
Access: Open weights
License: Tencent Hunyuan license
Rarity: Uncommon
Set number: 991/1044

Open Tencent docs Weights on Hugging Face Quality 88 Where to run it 20 Evolution 2/2 Race it 41 tok/s Meet its creature Battle Gemini 3.5 Flash Lite

No matching Hugging Face repo was found, so this shows Tencent’s usual open-weight license. Check the model card before relying on it. Specs from models.dev, downloads from Hugging Face, last changed Oct 19, 2026. Official source: cloud.tencent.com.

How good is it?

Public leaderboard scores for Hy4 preview, refreshed daily from Epoch AI and LMArena.

Top 12% of the 145 models on the LMArena WebDev: #19.

Quality is the share of a leaderboard’s models this one beats, from 0 to 100. Chat models are rated on Epoch AI’s Capabilities Index when it lists them, otherwise on LMArena; image and video models on LMArena’s image and video arenas.

Leaderboards

LMArena WebDev#19 of 145
Each model builds a web app from the same prompt; people pick the better one.1,631 rating

Scores from LMArena, licensed CC BY 4.0. Boards list models under their own names, so a close match is shown with the name it was scored under. Credits

Where to run it

18 providers sell Hy4 preview through an API, and 2 Hugging Face hosts serve it. Prices are per million tokens, cheapest first.

CheapestVancine$0.67 in · $2.00 out
From the labTencent TokenHub$0.83 in · $2.50 out
Fastest on Hugging FaceNovita41 tokens/s

API providers

Provider	Input	Output	Context
VancineCheapest paid`hy4-preview`	$0.67	$2.00	1.02M
CommandCode`tencent/hy4-preview`	$0.83	$2.50	1.02M
Deep Infra`tencent/Hy4-preview`	$0.83	$2.50	1.05M
DevPass (LLM Gateway)`hy4-preview`	$0.83	$2.50	1M
Eden AI`tencent/hy4-preview`	$0.83	$2.50	1M
Kilo Gateway`tencent/hy4-preview`	$0.83	$2.50	1.05M
LLM Gateway`tencent/hy4-preview`	$0.83	$2.50	1M
NanoGPT`tencent/hy4-preview`	$0.83	$2.50	1.05M
Novita AI`tencent/hy4-preview`	$0.83	$2.50	1M
OpenCode Go`hy4-preview`	$0.83	$2.50	1.02M

Show 8 more providers

Provider	Input	Output	Context
OpenRouter`tencent/hy4-preview`	$0.83	$2.50	1.05M
Tencent Token Plan`hy4-preview`	$0.83	$2.50	1.02M
Tencent TokenHubOfficial`hy4-preview`	$0.83	$2.50	1.02M
Vercel AI Gateway`tencent/hy4-preview`	$0.83	$2.50	1.02M
AIHubMix`hy4-preview`	$0.84	$2.54	1.02M
CrossModel`tencent/hy4-preview`	$0.96	$2.88	1.05M
Venice AI`hy4-preview`	$1.04	$3.13	1.05M
AI/ML API`tencent/hy4-preview`	$1.15	$3.44	1.05M

Hugging Face Inference Providers

Host	Input	Output	Context	Speed	First token
DeepInfraTools	$0.83	$2.50	1.05M	31 tok/s	1.3s
NovitaFastestTools	$0.83	$2.50	1M	41 tok/s	2.4s

From models.dev and Hugging Face, checked daily. Resellers’ listings can lag behind their price pages, so confirm with the provider before you commit.

Face off

Current text models from other labs at a similar output price.

Hy4 preview

Hy4 preview

How good is it?

Leaderboards

Where to run it

API providers

Hugging Face Inference Providers

Face off

More from Tencent

Hy-Memory-30B-A3B

AuK

Hy3

Hy3 preview

HunyuanOCR

Hunyuan3D-2