Model recipes are organized in three levels:
- Choose a task using the Hugging Face task taxonomy below.
- Open the task to see every model family with a declared recipe.
- Open a family to see its exact manifest recipes, applicable
trtmc CLI
commands, Hugging Face model_type and architecture or pipeline class,
TRTMC task/head contract, and every family-owned --set configuration key.
Every recipe row is generated from a model-owned manifest under
tests/e2e/models/<family>/manifests/. A manifest declares an executable test
contract; it is not, by itself, a current hardware pass receipt. Family config
tables are generated from registered Python or C++ config schemas rather than
maintained by hand. Use
Supported Models for the retained release-support snapshot.
Multimodal
| Hugging Face task | Model families | Declared recipes | What it covers |
|---|
| Any-to-Any | 1 | 1 | Models that accept or produce more than one modality through one task contract. |
| Image-Text-to-Text | 6 | 14 | Vision-language models that answer or generate text from an image and prompt. |
Natural Language Processing
| Hugging Face task | Model families | Declared recipes | What it covers |
|---|
| Text Generation | 34 | 88 | Decoder, encoder-decoder, recurrent, and diffusion-style text generation. |
| Translation | 3 | 4 | Text generation recipes whose declared user contract is translation. |
| Feature Extraction | 13 | 35 | Encoder and embedding models that return vector representations. |
| Text Ranking | 1 | 2 | Models that score documents against a query. |
Computer Vision
| Hugging Face task | Model families | Declared recipes | What it covers |
|---|
| Image Classification | 1 | 2 | Models that assign a class to an input image. |
| Image Segmentation | 1 | 2 | Models that produce segmentation masks for an image. |
| Mask Generation | 2 | 3 | Prompted segmentation models that generate masks from points or text. |
| Text-to-Image | 4 | 20 | Diffusion recipes that generate an image from text. |
| Image-to-Image | 1 | 1 | Image-conditioned generation and editing recipes. |
| Text-to-Video | 4 | 8 | Diffusion recipes that generate video frames from text. |
| Image-to-Video | 1 | 1 | Image-conditioned video and world-model generation recipes. |
Audio
| Hugging Face task | Model families | Declared recipes | What it covers |
|---|
| Automatic Speech Recognition | 3 | 9 | Speech-to-text models, including offline and streaming contracts. |
| Text-to-Speech | 2 | 7 | Models that synthesize audio from text. |
| Audio-to-Audio | 1 | 3 | Speech-to-speech models that consume and generate audio. |
Time Series
| Hugging Face task | Model families | Declared recipes | What it covers |
|---|
| Time Series Forecasting | 4 | 9 | Forecasting and neural-operator models that consume numerical sequences. |