# GSt APIs

Source: [https://docs.qualcomm.com/doc/80-70023-50/topic/gst-apis.html](https://docs.qualcomm.com/doc/80-70023-50/topic/gst-apis.html)

These APIs provide the required custom classes and layers to use hardware
        acceleration in the Qualcomm IM SDK and enable the tensor-based method/workflow for
        ML.

The following table lists all the supported APIs:

Table : GSt video and ML APIs

| API | Description |
| :--- | :--- |
| **Video APIs** |
| GstImageBufferPool APIs | Special GstBufferPool subclass to configure raw video<br>                            buffers. |
| GstGlesVideoConverter APIs | Use the GPU through the OpenGL ES based IB2C library to provide the<br>                            following: <ul class="ul" id="gst-apis__ul_d22_bkg_tcc"><br>                                <li class="li">Hardware-accelerated image manipulation</li><br><br>                                <li class="li">Transformation</li><br><br>                                <li class="li">Color conversion operations</li><br><br>                            </ul> |
| **Machine learning APIs** |
| GstMLType | Gets information on the GStreamer ML data types. |
| GstMLInfo | Represents the information structure describing the machine learning<br>                            properties. |
| GstMLTensorMeta | Defines the metadata API for different tensor types. |
| GstMLFrame | Contains mapped pointers to the tensors and the required information<br>                            about them. |
| GstMLBufferPool | Special GstBufferPool subclass to configure ML tensor<br>                            buffers. |

- **Video APIs**  

Use the video APIs to configure the video buffers and video converter.
- **Machine learning APIs**  

Configure the tensor-specific requirements and machine learning         properties.

**Parent Topic:** Use GStreamer utilities for custom plugins

Last Published: Mar 27, 2026

Previous Topic
 
Use GStreamer utilities for custom plugins Next Topic

Video APIs