ScaledThought
← All models

Tencent

Hy4 preview

Live · US + CAtextApache 2.0

Tencent's largest open-weight MoE yet, aimed at coding, office and scientific workflows.

Price per 1M tokens

Input

$0.834

Cached input

$0.042

Output

$2.50

Priced per model · Same in US and CA · Free tier included · Pricing

Specs

Model id
tencent-hy4-preview
Weights released
Aug 28, 2026
Parameters
770B total · 49B active
Architecture
MoE
Context
1M tokens (1,000,000)
Input
text
License
Apache 2.0
Regions
us, ca

Benchmarks

Independent scores aren't published yet. They'll appear here as soon as they are.

Sources: huggingface.co · github.com · tencent.com

Use it

Three lines of code, or zero.

OpenAI-compatible. Use the model id below, or auto to let the router decide when this model is worth it.

Quickstart→
import os
from openai import OpenAI

client = OpenAI(base_url=os.environ["ST_BASE_URL"], api_key=os.environ["ST_API_KEY"])
resp = client.chat.completions.create(
    model="tencent-hy4-preview",
    messages=[{"role": "user", "content": "Hello"}],
)

Opens your agent with the prompt typed. You press Enter.