Skip to content

Tencent · Chat

Hy 4

Tencent 🇨🇳ChatHy4 familyOne of its sources says the weights aren't publicly available.
From Aiml API's listing:
Hy4 is the official release of Tencent’s Hy4 model, succeeding Hy4 preview with stronger agentic capabilities and better performance on long-horizon, real-world productivity tasks. It accepts text and image input, has a 1M-token context window and configurable reasoning effort.
llminfo IQ
137.4
Max context
1M
Listings
38

Evidence and sources

🤗 tencent/Hy4llminfo IQ 137.4, ranked among chat models. From 46 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 123.4, 95% interval 111–136. From 3 benchmark families across 3 sources. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (56.3). Cheaper per task than its benchmark results predict. From 12 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsChatty (61 ±9). Generates more tokens than models with similar coding results. Measured at max effort. Little data behind it, so treat it as a rough estimate.Median 368,811 output tokens per task at max effort.How the Yap score worksReleased 28 Aug 2026Maker HQ: CN (according to Tencent's own website)
All rankings: #52 on OpenRouter today #53 on OpenRouter this week #54 on OpenRouter this month #256 on NanoGPT recentlySources: LMArena data, CometAPI Pricing data, Models.dev data, AI/ML API data, llminfo researchPrimary data from LMArena from CometAPI Pricing: name from Models.dev: weights availability from AI/ML API: primary mode, description from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Trade-offs

No open weights

The weights aren't public, so you can't run it yourself.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.

Input & output

as stated by the model's providers
Text
Images
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt caching

Listings

 Tencenttencent/hy4 · fp4Source: LiteLLM data960K$0.417cache read $0.021$1.2505
 Nous Portaltencent/hy4Source: Nous Portal Inference Gateway data1M$0.1668cache read $0.0084$0.5002
 Kenarihy4Source: Models.dev data1M$0.3429cache read $0.0343$1.20
 Atlascloudtencent/hy4Source: AtlasCloud data1M$0.417cache read $0.021$1.2505
 Hugging Facetencent/Hy4Source: Models.dev data1M$0.417$1.2505
 NanoGPTtencent/hy4Sources: Models.dev data, NanoGPT1M$0.417cache read $0.021$1.2505
 OpenRoutertencent/hy4Sources: Models.dev data, OpenRouter data docs1M$0.417cache read $0.021$1.2505
 Verceltencent/hy4Sources: Models.dev data, Vercel AI Gateway data1M$0.417cache read $0.021$1.2505
 Gmicloudtencent/hy4Sources: GMI Cloud data, LiteLLM data1M$0.417cache read $0.021$1.2505
 Novita AItencent/hy4Sources: Models.dev data, Novita AI data1M$0.417cache read $0.021$1.2505
Show all 35 listings

Showing 10 of 35: the maker's own listing first, then the cheapest per 1M input tokens from each provider.

Also sold by Aihubmix, Chutes, Cometapi, Crossmodel, Eden AI, Flowdot, Kilo, Llmgateway, Llmgateway Providers, Nan, Opencode Go, Ppio, Qiniu, Together AI, Wavespeed, Zenmux.

3 listings without a price are hidden.

Included in plans

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Agentic

Arena AgentSource: LMArena data · Nov 2026
ValsSource: Vals AI

Coding

Arena ELO WebdevSource: LMArena data · Nov 2026

Human preference

Design Arena EloSource: Design Arena Leaderboards data · Nov 2026

Reasoning

General

Science

Other

Artificial AnalysisSource: OpenRouter data docs
SuperCLUESource: SuperCLUE data
ValsSource: Vals AI

Data credits: LMArena, CC BY 4.0 (licence) · BullshitBench by Peter Gostev, MIT (licence) · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence)

Also known as: gmi/tencent/hy4, tencent/hy4, together-ai/hy4, together_ai/tencent/hy4

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works