Tencent

Tencent Hy4 Flash

The smaller Hy4 model for frequent coding, document and agent tasks, with optional reasoning and room for a large set of sources.

224B total · 12B activetext inputtext output

Run this model

tencent/hy4-flash

Use an endpoint supported by this model.

Get a key for this setup →
Thinking guide →
const response = await fetch('https://api.minirouter.sh/v1/chat/completions', {  method: 'POST',  headers: {    Authorization: `Bearer ${process.env.MINIROUTER_KEY}`,    'Content-Type': 'application/json',  },  body: JSON.stringify({    model: 'tencent/hy4-flash',    messages: [{ role: 'user', content: 'Why is the sky blue?' }],    max_tokens: 1024,  }),}) const data = await response.json()console.log(data.choices[0].message.content)
Read docs →

Text input only. Explicit reasoning settings can reserve the full output ceiling before the request; the final charge follows actual usage.

Model overview

Input / 1M
$0.03675
All-in MiniRouter rate
Output / 1M
$0.147525
All-in MiniRouter rate
Context
1M
1,024,000 tokens
Max output
64K
tokens
Reasoning
none / high
Optional thinking
Best fit
Everyday work
Coding assistance and document processing
Announced
16 Nov 2026
publisher release

Compare text-model alternatives or estimate token costs for your workload.

Practical use

Use Flash for the repeatable steps

A large context window can keep related material together. A narrow instruction and a clear check still make each answer easier to trust.
  1. [ 01 ]

    Prepare the text

    Collect the relevant passages, code and task instructions. Give each source a label so the answer can point back to it.

  2. [ 02 ]

    Draft with Flash

    Ask for a focused summary, a proposed edit or an explanation. Keep each step small enough to check independently.

  3. [ 03 ]

    Check and escalate

    Compare claims with the sources and run tests for code. Move unresolved, long-running engineering work to a stronger model.

Tencent's design

A smaller active model, a large working context

Tencent reports 224 billion total parameters, including 64 billion Engram parameters, with 12 billion active. Flash is a lower-cost sibling of Hy4, not a replacement for its hardest workloads.

Model and endpoint specifications

Active parameters
12BTencent's model specification
Context tokens
1,024,000Published gateway limit
Maximum output
64,000Published gateway ceiling
Tencent announcement ↗ · Gateway endpoint specifications ↗. Read November 16, 2026.
View chart data
Hy4 Flash specifications
SpecificationValue
Total parameters224B, including 64B Engram
Active parameters12B
Context tokens1,024,000
Maximum output tokens64,000
Before you switch

Keep the task and the route in view

Use the current catalog prices above to budget your workload. Token prices alone do not tell you the cost of a successfully completed task.
  1. [ 01 ]

    Start with text

    Paste extracted document text or code. Image inputs and schema-constrained outputs are not available on this MiniRouter route.

  2. [ 02 ]

    Compare on your own work

    Try a small set of representative tasks and inspect correctness, retries and total usage. Publisher evaluations do not measure your application.

  3. [ 03 ]

    Keep Hy4 for demanding work

    Tencent positions the larger Hy4 for complex, long-running software engineering. Flash targets the frequent steps around it.

Route

One route, standard clients.

Authenticate with a MiniRouter bearer key. Caller-provided upstream credentials are not used on this route.
Base URLhttps://api.minirouter.sh/v1
Model IDtencent/hy4-flash
Endpoints/v1/chat/completions/v1/messages/v1/responses
Parametersmax_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoning
Structured outputNot enforced; a requested format is dropped and the answer is plain text.
ModalitiesInput: text · Output: text

Configured providers. Priority follows the live catalog order.

ProviderPriorityInput / 1MOutput / 1MParametersAccess
vercel
1$0.03675$0.147525max_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoningMiniRouter balance

Every model, one key.