Hy4
Hy4 is a large language model made by Tencent, added to the catalog on 4 Nov 2026. It has a context window of 1,048,576 tokens and costs $0.417 per 1M input tokens and $1.251 per 1M output tokens. It accepts text and images and produces text. Typical uses: interpreting images with accompanying text, analyzing long documents and tool-assisted productivity workflows.
Price checked ยท source: OpenRouter
What Hy4 is
Tencent Hy4 accepts text and images and returns text. Its listed inputs allow prompts that combine written instructions with visual material, as well as text-only requests. The catalog gives it a context window of 1,048,576 tokens and a maximum output of 65,536 tokens. That context capacity can accommodate extensive supplied material, while the output limit allows long responses. The listed modalities and limits describe the interaction format; the facts do not specify image formats or particular image-analysis tasks.
Hy4 also supports reasoning, tool use, structured outputs, and response-format controls. These capabilities can suit applications that need to request a particular output format or connect the model to tools. The catalog does not name specific tools or schemas, so what is available depends on the application using the model. Tencent describes the official release as improving on Hy4 Preview, with gains in agentic capabilities and performance across long-horizon, real-world productivity work. The supplied facts do not include measurements or task-by-task results. Its large context window may suit analyzing long documents or working with extensive reference material, while image input allows visual material to be included in a prompt. Tool support may fit workflows that call tools, and structured outputs may suit tasks that need text in a specified format. These are potential uses based on the listed capabilities, not guarantees of performance.
Key features
- Accepts text and image inputs and produces text
- Supports a context window of 1,048,576 tokens
- Allows outputs up to 65,536 tokens
- Supports tools, reasoning, structured outputs, and response formats
Typical uses
- Interpreting images with accompanying text
- Analyzing long documents
- Tool-assisted productivity workflows
- Generating structured text responses
Hy4 pricing
US dollars, list price through OpenRouter.
| Item | Standard |
|---|---|
| Input, per 1M tokens | $0.417 |
| Output, per 1M tokens | $1.251 |
| Cached input read, per 1M tokens | $0.021 |
No price change since it was added on 4 Nov 2026.
| What a task costs | Cost |
|---|---|
| Answer a short question500 in / 300 out | $0.0006 |
| Summarise a 50-page PDF25,000 in / 800 out | $0.011 |
| Review a 2,000-line source file30,000 in / 2,000 out | $0.015 |
| Read a 300-page contract set150,000 in / 1,500 out | $0.064 |
Hy4 specifications
| Model id | tencent/hy4 |
|---|---|
| Version | hy4-20261104 |
| Context window | 1,048,576 tokensAbout 786,432 English words |
| Max output | 65,536 tokens |
| Knowledge cutoff | Not published |
| Added to catalog | 4 Nov 2026 |
| Input types | text, image |
|---|---|
| Output types | text |
| Tool calling | Yes |
| Structured outputs | Yes |
| Reasoning effort control | Yes |
| Content moderation | Not applied by provider |
Sample outputs
The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.
No output has been recorded for Hy4 yet.
Compare Hy4
Side by side with other models: prices, context window, what each accepts and supports, and the same four tasks priced on both.