hy4 Preview API

Tencent's hy4 Preview, through an OpenAI-compatible API. Providers quote it on an open exchange, and each request goes to the offer with the lowest estimated charge that meets your rules.

Providers compete for each request. The exchange ranks the offers that meet your rules by estimated charge. It fixes your maximum charge before the first token, and you pay for the tokens that the model uses.
The OpenAI and Anthropic APIs. Use the OpenAI or Anthropic SDK that you have. Set the base URL and your key, and send hy4-preview as the model.
Your data rules. Send x-liquid-retention: none and only offers whose provider does not keep your content compete for the request.

Facts

Model name: hy4-preview
Maker: Tencent
Context window: 991.8K tokens
Input price: $1.042 per million tokens
Output price: $3.126 per million tokens
Providers: 1 independent provider
Regions: us

Providers

Pera: us, $1.042 input and $3.126 output per million tokens, $1.042 cache read per million tokens, 991.8K tokens of context

Scores

HLE: 27.8, #88 of 273
GPQA: 86.7, #62 of 250
AA-LCR: 64.7, #135 of 237
Terminal-Bench Hard: 34.1, #45 of 166
τ²-Bench: 92.7, #30 of 169
IFBench: 63.1, #60 of 178
SimpleQA Verified: 52.0, #27 of 126

Scores: Artificial Analysis (https://artificialanalysis.ai/)

Scores: Epoch AI, 'Capabilities & Benchmarking' (https://epoch.ai/benchmarks)

Request

curl https://router.inference.ai.exchange/v1/chat/completions \
  -H "Authorization: Bearer $LIQUID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "hy4-preview",
    "messages": [{"role":"user","content":"Say hello."}]
  }'

Questions

What is hy4 Preview?: hy4 Preview is a model from Tencent.
How much does hy4 Preview cost?: Offers now ask $1.042 per million input tokens and $3.126 per million output tokens. A job pays for the tokens it uses. The exchange picks the cheapest eligible offer.
What is the context length of hy4 Preview?: The largest offer takes 991,808 input tokens. One job can generate up to 131,072 tokens. A smaller offer can take less. The exchange skips an offer that cannot take your message.
What can hy4 Preview do?: Its providers list Reasoning effort, JSON object, JSON schema, Frequency penalty, Presence penalty, Seed, Stop sequences, Top-p sampling, Text, Tools.
What inputs and outputs does hy4 Preview support?: Text → Text.
Which providers serve hy4 Preview?: 1 independent provider quote it now, with 1 offer. The table above ranks them by the cost of a typical job.
How do I call hy4 Preview?: Send "hy4-preview" as the model in an OpenAI or Anthropic request. The API section links to setup steps for each client.
When was hy4 Preview listed?: The exchange first listed it on 9/29/2026. This is not its release date.