Skip to content

Image Generation Model Keys

Currently, WiseMind AI requires users to configure their own image generation model API keys. The following service providers are supported:

All costs incurred by using these keys are charged directly by the respective model provider and have no relation to WiseMind AI.

Service ProviderApplication AddressOfficial Website
Agnes AIGet API KeyOfficial Website
Alibaba Cloud Bailian (阿里云百炼)Get API KeyOfficial Website
ByteDance Doubao (字节豆包)Get API KeyOfficial Website
GeminiGet API KeyOfficial Website
MiniMax AIGet API KeyOfficial Website
ModelScope (魔搭社区)Get Access TokenOfficial Website
OllamaNot required (Local Service)Official Website
OpenAIGet API KeyOfficial Website
OpenRouterGet API KeyOfficial Website
Tongyi Qianwen (通义千问)Get API KeysOfficial Website
SiliconCloudGet Access KeyOfficial Website
Zhipu AI (智谱 AI)Get API KeysOfficial Website
Tencent Hunyuan TokenHubGet API KeyOfficial Website
CloudflareGet API TokenModel List
StepFunGet API KeyOfficial Website

INFO

More image generation services are being added soon. Stay tuned!

1. How to Apply for a ModelScope (魔搭社区) API Token

First, open the ModelScope token application page. If you haven't registered yet, please do so first. Then follow these steps:

  1. Log in and navigate to the Access Tokens page to create or copy an existing token;
  2. Open WiseMind AI's SettingsModel Settings, then locate Image Generation Model;
  3. Select ModelScope (魔搭社区), paste your copied token, and save.

Once completed, you can use ModelScope's text-to-image models in WiseMindAI.

2. Configure the new image model providers

Tencent Hunyuan TokenHub

Create an API key in Tencent Cloud TokenHub. In WiseMindAI, open Settings → Model Settings → Image Generation Models, select Tencent Hunyuan TokenHub, and enter the API key. The endpoint can usually remain at its default value.

Cloudflare

Cloudflare image generation requires:

  • Account ID
  • API Token
  • Image model name

The API token must have permission to call Workers AI. Choose a model name from the Cloudflare Workers AI image generation model list.

StepFun

Create an API key on the StepFun platform. Select StepFun in WiseMindAI, then enter the key, model name, and endpoint.

3. Test an image model directly

After configuring an online image model, select Test. WiseMindAI uses a built-in prompt to generate a test image, confirming that the API key, model name, and endpoint work correctly.

Testing does not change the active image model or save the result to a production document. It makes a real service request and may incur a small fee.

Recommendations

  • Generate a test image after the initial setup before using Knowledge Infographics or generating illustrations for notes.
  • Supported sizes, aspect ratios, and parameters vary by provider. Review the form after switching providers.
  • If the model is reported as unavailable, confirm its exact name and your account permissions with the provider.
  • If an image is generated but contains too much text or an unclear structure, organize the draft first in Knowledge Infographic, then generate the final image.