Author's Description
Hy3 preview is a high-efficiency Mixture-of-Experts model from Tencent designed for agentic workflows and production use. It supports configurable reasoning levels across disabled, low, and high modes, allowing it to...
Key Specifications
Supported Parameters
This model supports the following parameters:
Features
This model supports the following features:
Performance Summary
Tencent's Hy3 preview, a high-efficiency Mixture-of-Experts model, demonstrates strong performance in several key areas, particularly for agentic workflows. While its speed ranking indicates it tends to have longer response times (17th percentile), its pricing is competitive (44th percentile). A standout feature is its exceptional reliability, boasting a 97% success rate across benchmarks. The model achieves perfect accuracy in General Knowledge and Ethics, with the latter also being the most accurate among models at its price point and speed. It shows strong capabilities in Reasoning (98% accuracy, 92nd percentile) and Instruction Following (79% accuracy, 86th percentile), indicating its suitability for complex tasks. Coding and Mathematics also show solid performance at 93% and 94% accuracy respectively. A notable weakness is its performance in Hallucinations, where it scored 84% accuracy (26th percentile), suggesting room for improvement in acknowledging uncertainty. Overall, Hy3 preview is a reliable and accurate model, particularly strong in knowledge-based and reasoning tasks, though its speed and hallucination control could be enhanced.
Model Pricing
Current Pricing
| Feature | Price (per 1M tokens) |
|---|---|
| Prompt | $0.066 |
| Completion | $0.26 |
| Input Cache Read | $0.029 |
Price History
Available Endpoints
| Provider | Endpoint Name | Context Length | Pricing (Input) | Pricing (Output) |
|---|---|---|---|---|
|
SiliconFlow
|
SiliconFlow | tencent/hy3-preview-20260421 | 262K | $0.066 / 1M tokens | $0.26 / 1M tokens |
Benchmark Results
| Benchmark | Category | Reasoning | Strategy | Free | Executions | Accuracy | Cost | Duration |
|---|
Other Models by tencent
|
|
Released | Params | Context |
|
Speed | Ability | Cost |
|---|---|---|---|---|---|---|---|
| Tencent: Hunyuan A13B Instruct | Jul 08, 2025 | 13B | 131K |
Text input
Text output
|
★ | ★ | $$ |