Hy4 preview

tencent/hy4-preview
FlagshipFeatured
ToolsJSONReasoning
by Tencent · 2026-08-28

Hy4 preview is Tencent latest Mixture-of-Experts flagship: 770B total parameters with 49B activated per token, 78 layers, 256 routed experts plus 1 shared expert, and a 1M-token context window. It is a text-only model built for coding agents, complex tool-use workflows, and productivity work, with a native MTP layer for speculative decoding. Reasoning is configurable at three levels (high, low, none) and defaults to high, so it can run as a deep chain-of-thought model for math, coding and analysis, or answer directly when latency matters. It supports native tool calling, structured outputs, and the full sampling parameter set; Tencent recommends temperature 0.9 with top_p 1.0. Tencent positions it for software engineering, office and data analysis, game prototyping, and scientific research, and reports an internal blind evaluation (163 experts, 203 engineering tasks) in which Hy4 preview averaged 2.99 against GLM 5.3 at 2.92 and Kimi K3 at 2.94. As a preview release Tencent notes known rough edges, including spending longer than necessary on reasoning and over-verifying its own work.

ctx1M tokens
Max output64K
Inputtext
Outputtext
Best forreasoning, coding, agentic
p50 TTFT3.17 s
INPUT$0.83/ 1M tokens
OUTPUT$2.50/ 1M tokens
p50 TTFT3.17 s7d
p95 TTFT3.17 s7d
TRAFFIC38tokens / 7d

Code samples

Call from any SDK

OpenAI-compatible — keep the SDK you already use

  • OpenAI SDKhttps://api.orcarouter.ai/v1
import os

from openai import OpenAI

client = OpenAI(
    base_url="https://api.orcarouter.ai/v1",
    api_key=os.environ["ORCAROUTER_API_KEY"],
)

response = client.chat.completions.create(
    model="tencent/hy4-preview",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Supported parameters

  • frequency_penalty
  • include_reasoning
  • max_completion_tokens
  • max_tokens
  • presence_penalty
  • reasoning
  • reasoning_effort
  • repetition_penalty
  • response_format
  • seed
  • stop
  • structured_outputs
  • temperature
  • tool_choice
  • tools
  • top_k
  • top_p

Pricing

Pricing
Input / 1M tokens$0.834
Output / 1M tokens$2.501
Cache read / 1M$0.042
CurrencyUSD

Cost calculator

Tokens / month10MM
Input share70%%
Estimated / month $13.34 · With prompt caching $10.57

Estimate based on list price

Token & cost estimator

Input tokens: 20Cost per request: $0.001267

Estimate only — actual token counts depend on the provider's tokenizer.

Performance

p50 TTFT
3.17 s
Output speed
Collecting…
p95 TTFT
3.17 s
Error rate
0%

Public benchmarks

Source: Design Arena

Community buzz

What developers are saying this week

Hacker News0 mentions · 7d

How it compares

Hy4 previewHy3
Input $/M$0.83$0.18
Output $/M$2.50$0.59
Context1.0M262K
Quality8/108/10
Compare side-by-sideCompare side-by-side

FAQ

How much does Tencent: Hy4 preview cost on OrcaRouter?
Tencent: Hy4 preview is priced at $0.83 per 1M input tokens and $2.50 per 1M output tokens via OrcaRouter. Pricing is pulled live from the routing layer.
What is Tencent: Hy4 preview's context window?
Tencent: Hy4 preview supports a context window of 1M tokens. Use long-context features (RAG, summarisation) up to that limit.
How do I call Tencent: Hy4 preview via the OpenAI SDK?
Set OpenAI base_url to https://api.orcarouter.ai/v1, supply your OrcaRouter API key, and pass model="tencent/hy4-preview" in the chat.completions.create call.
Does OrcaRouter rate-limit Tencent: Hy4 preview?
Per-model rate limits follow your OrcaRouter plan. Free tiers ship with conservative caps; paid tiers lift them. Check /pricing for current quotas.

Embed this badge

Tencent: Hy4 preview•$0.83/M in•3175ms p50•via OrcaRouter
HTML <a href="https://www.orcarouter.ai/models/tencent/hy4-preview" target="_blank"> <img src="https://www.orcarouter.ai/embed/tencent/hy4-preview.svg" alt="Tencent: Hy4 preview on OrcaRouter" /> </a>
Markdown [![Tencent: Hy4 preview](https://www.orcarouter.ai/embed/tencent/hy4-preview.svg)](https://www.orcarouter.ai/models/tencent/hy4-preview)

Model card as data

GET /api/public/models/tencent/hy4-previewOpen
Machine-readable:/llms.txt/llms-full.txt