Author: WaveSpeedAI Version: 0.0.2 Type: tool Source: https://github.com/WaveSpeedAI/dify-wavespeed-plugin Contact: support@wavespeed.ai
Generate images and videos in Dify with WaveSpeed AI — a fast inference platform serving state-of-the-art media generation models such as ByteDance Seedream (image) and Seedance (video), plus hundreds of other models.
- Create an account at wavespeed.ai.
- Get your API key from the API Keys page of the dashboard.
- Install this plugin in Dify, open its authorization settings, and paste the key into WaveSpeed API Key. The plugin validates the key with a lightweight authenticated call when you save.
The plugin makes outbound HTTPS requests to a single fixed endpoint: https://api.wavespeed.ai. Authentication uses your API key as a Bearer token. No other network destinations are contacted, and no inbound connections are needed.
Generate an image from a text prompt.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
prompt |
string | yes | — | Text description of the image |
model |
string | no | bytedance/seedream-v5.0-pro |
WaveSpeed model id |
size |
string | no | model default | Resolution as width*height, e.g. 2048*2048 |
seed |
number | no | random | Seed for reproducible results |
Returns the generated image(s) plus a JSON payload with the task id and output URLs.
Generate a video from a text prompt.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
prompt |
string | yes | — | Text description of the video |
model |
string | no | bytedance/seedance-2.5/text-to-video |
WaveSpeed model id |
duration |
number | no | model default | Video duration in seconds (supported values depend on the model) |
Returns the generated video URL as a link and text message, plus a JSON payload with the task id and output URLs.
Use either tool in an Agent, Chatflow, or Workflow application. In a workflow, add the tool node, wire the prompt from an upstream variable, and optionally set model, size/duration, or seed. Browse the full model catalog at wavespeed.ai/models — any compatible image or video model id can be passed in model.
Generation is synchronous from Dify's point of view: the tool submits the task and polls the WaveSpeed API (every second, up to 10 minutes) until the media is ready.
See PRIVACY.md. The plugin stores nothing; prompts and parameters are sent only to the WaveSpeed API.
- Email: support@wavespeed.ai
- Website: https://wavespeed.ai
- Source repository: https://github.com/WaveSpeedAI/dify-wavespeed-plugin