hy3

Live
tencent

Hy3 is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team. Following the Hy3 Preview launch in late April, we gathered feedback from 50+ products and scaled up post-training with higher quality data. Today, we introduce Hy3, which outperforms similar-size models and rivals flagship open-source models with 2-5x parameters. It also shows significant gains in utility across various products and productivity tasks.

Modality

Chat

Context window

262K tokens

Max output

16K tokens

Region

US

Published

Jul 21, 2026

Pricing

Input

7.87

EGP per 1M tokens

Output

32.60

EGP per 1M tokens

Use it via the API

hy3 works with any OpenAI-compatible SDK — point the base URL at https://preprod-backend.sovereigneg.com/v1 and use your SovereignEG API key.

Run inference

OpenAI-compatible — POST /v1/chat/completions — drop-in for any OpenAI SDK.

from openai import OpenAI

client = OpenAI(
    base_url="https://preprod-backend.sovereigneg.com/v1",
    api_key="YOUR_API_KEY",
)

response = client.chat.completions.create(
    model="hy3",
    messages=[
        {"role": "user", "content": "What are some fun things to do in Cairo?"}
    ],
)

print(response.choices[0].message.content)

Frequently asked questions

How much does hy3 cost?

7.87 EGP per 1M input tokens; 32.60 EGP per 1M output tokens. Billing is metered per token in Egyptian pounds (EGP), with no minimum commitment.

What is the context window of hy3?

hy3 supports a context window of 262,144 tokens (262K), with up to 16,384 output tokens per response.

Where is hy3 hosted?

hy3 is served from US (US-hosted inference).

How do I use hy3 via the API?

hy3 is available through the OpenAI-compatible SovereignEG API: point your SDK's base URL at https://preprod-backend.sovereigneg.com/v1 and call /v1/chat/completions with model "hy3" and your API key.