Introduction¶

A common approach to adding video generation capabilities to an application is to directly integrate with a model provider’s native API. However, these interfaces vary in parameters, often requiring code changes when switching models or relay stations. Another approach is to connect to an OpenAI-compatible relay station, using a unified request format to call different backend models.

The following introduces a Skill that follows this route: the Video Generation Skill. It calls the Doubao Seedance video generation API through the Thingo relay station (OpenAI-compatible format), converging text-to-video, first-frame image-to-video, and video extension/editing into a single set of interfaces.

What is this¶

The Video Generation Skill is a video generation skill capable of connecting to relay stations, compatible with OpenAI interface calls. It is version 1.0.0, sourced from the community, and categorized under design-media.

The publisher is located at @user_1707615b/t-h-i-n-g-o-2026video, and the official address is https://api.skillhub.cn/user_1707615b/t-h-i-n-g-o-2026video. It solves the problem of “one Skill covering multiple video generation tasks”: without writing multiple sets of integration logic, it uses the same set of OpenAI-compatible requests to accomplish text-to-video, image-to-video, and video extension/editing respectively.

Core Features¶

1. Text-to-Video: Generate videos from input prompts.
2. First-frame Image-to-Video: Generate subsequent video based on a given first-frame image.
3. Video Extension/Editing: Extend or edit existing videos.
4. Aspect Ratio Support: Supports multiple ratios such as 16:9, 9:16, and 1:1.

The source high-quality model address is https://uniapi.thingo.com.cn (Thingo Large Model Aggregation Platform). The Skill adopts the general SKILL.md format and is marked with requires_api_key=true, meaning an API Key is required for invocation.

Installation and Activation¶

According to the scraped data, this Skill is provided in the general SKILL.md format, sourced from the community, and marked with requires_api_key=true, meaning the relay station’s API Key needs to be configured before activation.

The data does not provide directly verifiable official installation commands, directory paths, or configuration fields. Without verifiable commands, installation steps should not be fabricated. It is recommended to view the Skill’s SKILL.md and installation instructions via the official address https://api.skillhub.cn/user_1707615b/t-h-i-n-g-o-2026video, and confirm the required environment variable names, interface addresses, and authentication fields before enabling it.

Typical Usage¶

The data does not provide field-by-field request examples. The following is the reproducible calling direction based on its description: Send requests to the Thingo relay station via the OpenAI-compatible format, selecting text-to-video, first-frame image-to-video, or video extension/editing as needed; specify the aspect ratio among 16:9, 9:16, or 1:1. Please refer to the official SKILL.md for specific paths and parameters.

Applicable Scenarios and Notes¶

Suitable for scenarios where video generation needs to be integrated into an application and unified connection via OpenAI-compatible format is desired; also suitable for teams that need to switch between text-to-video, first-frame image-to-video, and video extension/editing.

Note the following points:

1. An API Key is required (requires_api_key=true); it cannot be called without configuration.
2. Capabilities depend on the relay station backend; the final available models and parameters are subject to the Thingo relay station.
3. The official page is currently inaccessible; ratings, downloads, and other data are based on the time of scraping (Rating 100000, Downloads 1879, Installs 0).

Conclusion¶

The value of this Skill lies in converging multiple video generation tasks into a single set of OpenAI-compatible interfaces, reducing integration costs. Official address: https://api.skillhub.cn/user_1707615b/t-h-i-n-g-o-2026video.