DeepSeek-v4.1-flash Joins WeChat Mini-Program’s 10-Billion-Token Free Tier
The AI resource pack granted by the Mini-Program Growth Plan has two new deductible models today: hy4-preview and deepseek-v4.1-flash. Together with the existing hy3, the 10 billion Token text-generation allowance now supports three text models.
With this update, none of the three models need to be enabled in the Cloud Development console. When calls are made from the mini-program, Cloud Development server, or Cloud Development console, Token usage is automatically deducted from the granted allowance; no extra configuration is needed.
▍How to use it in your mini-program
You can directly use the table below to apply the new capabilities to your business:
| Recommended business scenarios | What to change |
|---|---|
| Photo recognition (photo receipt one-click reimbursement, test paper auto-grading, product screenshot auto-categorization, image table extraction) | Replace model with "deepseek-v4.1-flash" and pass the image's cloud storage URL directly. |
| Long text processing (batch parsing of long documents, summarization of long customer-service chat logs, multi-turn long conversations for intelligent assistants) | Replace model with "hy4-preview"; no need to rely on complicated context cropping or truncation logic. |
The image generation model Hy-image-v3.5-preview and the just-released formal Hy4 are also being integrated, and will be announced after launch.
▍How to claim
All mini-program developers can apply. Each mini-program can receive a free Cloud Development environment, a 10-billion-Token text-generation allowance, and a 100,000-image generation allowance per activity rules. It now supports designated models such as Hy4 Preview and DeepSeek. Welcome to experience it in the Mini-Program Growth Plan.
Application entry: WeChat Public Platform → Industry Capabilities → Mini-Program Growth Plan.
What kind of mini-program are you building? Or do you have any questions during development? Feel free to discuss in the comments～
Related documentation: