Requesty
Novita AI logo
Novita AI

tencent/hy3

Hy3 is built for real world business scenarios with a 295B total, 21B active MoE architecture, native 256K context support, and three reasoning modes. It enhances coding, long form comprehension, multi turn dialogue, and agentic task execution, balancing reliability, efficiency, and cost across high frequency interactions and complex workflows.

ReasoningTool calling
CompareDocs

Specifications

Context window262K tokens
Max output262K tokens
API typechat
AddedJul 13, 2026
Model IDnovita/tencent/hy3
Data retentionYes
Used for trainingUnknown
Provider location🇺🇸 US
Privacy policyNovita AI Privacy Policy→

Benchmarks

Released 2026-07-06
Coding Indexcoding
58.8%

Artificial Analysis Coding Index — a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
89.7%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
41.2%

Artificial Analysis Intelligence Index — a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality. Always test on your own workload.

Pricing

Input / 1M
Free
Output / 1M
Free
Cache write
N/A
Cache read
N/A
Estimated cost
100K input + 10K outputFree
1M input + 100K outputFree
10M input + 1M outputFree

Requesty charges exactly what the upstream provider charges, no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to novita/tencent/hy3.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="novita/tencent/hy3",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

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

Other Novita AI models

zai-org/glm-5.1
205K ctx$1.38/1M
Coding 56
xiaomimimo/mimo-v2-pro
1.0M ctx$2.00/1M
xiaomimimo/mimo-v2-flash
262K ctx$0.10/1M
minimax/minimax-m2.7-highspeed
205K ctx$0.60/1M
kwaipilot/kat-coder-pro
256K ctx$0.30/1M
inclusionai/ring-2.6-1t
262K ctx$0.30/1M
Coding 43

Frequently asked questions

How much does tencent/hy3 cost?
tencent/hy3 is priced at Free per million input tokens and Free per million output tokens when accessed via Requesty. Requesty charges exactly what the upstream provider charges, we don't add markup.
What is the context window of tencent/hy3?
tencent/hy3 has a context window of 262K tokens, with a maximum output of 262K tokens per response. That's roughly 350 words of input you can fit in a single prompt.
How does tencent/hy3 perform on benchmarks?
tencent/hy3 scores 89.7% on GPQA Diamond, 58.8% on Coding Index, 47.6% on SciCode. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can tencent/hy3 do?
tencent/hy3 supports tool calling, extended reasoning. You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use tencent/hy3 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "novita/tencent/hy3". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run tencent/hy3 through Requesty?
Yes. tencent/hy3 runs through Requesty's OpenAI-compatible API, served from Novita AI. You do not host the model yourself: point base_url at Requesty, set the model to "novita/tencent/hy3", and requests are routed to the upstream provider with automatic failover. The same key gives you 400+ other models too.

Access tencent/hy3 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

All Novita AI models