# Supported Models

mobius supports 298 registered model types.

## Text Generation

Standard autoregressive language models (CausalLM).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`DFlashDraftModel` | `DFlashDraftModel` | `dflash-draft` |
| {doc}`Eagle3DraftModel` | `Eagle3DraftModel` | `eagle3-draft` |
| {doc}`Eagle3LlamaForCausalLM` | `Eagle3DraftModel` | `eagle3-draft` |
| {doc}`Eagle3Speculator` | `Eagle3DraftModel` | `eagle3-draft` |
| {doc}`Gemma4AssistantForCausalLM` | `Gemma4AssistantCausalLMModel` | `gemma4-assistant` |
| {doc}`Gemma4UnifiedAssistantForCausalLM` | `Gemma4AssistantCausalLMModel` | `gemma4-assistant` |
| {doc}`LlamaForCausalLMEagle3` | `Eagle3DraftModel` | `eagle3-draft` |
| {doc}`Qwen35MtpModel` | `Qwen35MtpModel` | `qwen35-mtp` |
| {doc}`apertus` | `ApertusCausalLMModel` | `text-generation` |
| {doc}`arcee` | `ArceeCausalLMModel` | `text-generation` |
| {doc}`baichuan` | `CausalLMModel` | `text-generation` |
| {doc}`bloom` | `BloomCausalLMModel` | `text-generation` |
| {doc}`chatglm` | `ChatGLMCausalLMModel` | `text-generation` |
| {doc}`code_llama` | `CausalLMModel` | `text-generation` |
| {doc}`codegen` | `CodeGenCausalLMModel` | `text-generation` |
| {doc}`codegen2` | `CausalLMModel` | `text-generation` |
| {doc}`cohere` | `CohereCausalLMModel` | `text-generation` |
| {doc}`cohere2` | `CohereCausalLMModel` | `text-generation` |
| {doc}`command_r` | `CausalLMModel` | `text-generation` |
| {doc}`csm` | `CausalLMModel` | `text-generation` |
| {doc}`diffllama` | `DiffLlamaCausalLMModel` | `text-generation` |
| {doc}`doge` | `DogeCausalLMModel` | `text-generation` |
| {doc}`ernie4_5` | `ErnieCausalLMModel` | `text-generation` |
| {doc}`evolla` | `CausalLMModel` | `text-generation` |
| {doc}`exaone` | `CausalLMModel` | `text-generation` |
| {doc}`exaone4` | `ExaOne4CausalLMModel` | `text-generation` |
| {doc}`falcon` | `FalconCausalLMModel` | `text-generation` |
| {doc}`falcon_h1` | `FalconCausalLMModel` | `text-generation` |
| {doc}`falcon_mamba` | `MambaCausalLMModel` | `ssm-text-generation` |
| {doc}`gemma` | `GemmaCausalLMModel` | `text-generation` |
| {doc}`gemma2` | `Gemma2CausalLMModel` | `text-generation` |
| {doc}`gemma3_text` | `Gemma3CausalLMModel` | `text-generation` |
| {doc}`gemma3n` | `Gemma3nCausalLMModel` | `text-generation` |
| {doc}`gemma3n_text` | `Gemma3nCausalLMModel` | `text-generation` |
| {doc}`gemma4_assistant` | `Gemma4AssistantCausalLMModel` | `gemma4-assistant` |
| {doc}`gemma4_unified_assistant` | `Gemma4AssistantCausalLMModel` | `gemma4-assistant` |
| {doc}`glm` | `GlmCausalLMModel` | `text-generation` |
| {doc}`glm4` | `Glm4CausalLMModel` | `text-generation` |
| {doc}`glm4v_text` | `Glm4CausalLMModel` | `text-generation` |
| {doc}`gpt_neox` | `GPTNeoXCausalLMModel` | `text-generation` |
| {doc}`gpt_neox_japanese` | `GPTNeoXJapaneseCausalLMModel` | `text-generation` |
| {doc}`gptj` | `GPTJCausalLMModel` | `text-generation` |
| {doc}`granite` | `GraniteCausalLMModel` | `text-generation` |
| {doc}`helium` | `CausalLMModel` | `text-generation` |
| {doc}`hunyuan_v1_dense` | `HunYuanV1DenseCausalLMModel` | `text-generation` |
| {doc}`internlm2` | `InternLM2CausalLMModel` | `text-generation` |
| {doc}`llada` | `LLaDAModel` | `masked-diffusion` |
| {doc}`llama` | `CausalLMModel` | `text-generation` |
| {doc}`llama4_text` | `Llama4CausalLMModel` | `text-generation` |
| {doc}`mamba` | `MambaCausalLMModel` | `ssm-text-generation` |
| {doc}`mamba2` | `Mamba2CausalLMModel` | `ssm2-text-generation` |
| {doc}`minicpm` | `CausalLMModel` | `text-generation` |
| {doc}`minicpm3` | `CausalLMModel` | `text-generation` |
| {doc}`minimax` | `MiniMaxCausalLMModel` | `hybrid-text-generation` |
| {doc}`ministral` | `CausalLMModel` | `text-generation` |
| {doc}`ministral3` | `CausalLMModel` | `text-generation` |
| {doc}`mistral` | `CausalLMModel` | `text-generation` |
| {doc}`modernbert-decoder` | `ModernBertDecoderModel` | `text-generation` |
| {doc}`mpt` | `MPTCausalLMModel` | `text-generation` |
| {doc}`nanochat` | `NanoChatCausalLMModel` | `text-generation` |
| {doc}`nemotron` | `NemotronCausalLMModel` | `text-generation` |
| {doc}`olmo` | `OLMoCausalLMModel` | `text-generation` |
| {doc}`olmo2` | `OLMo2CausalLMModel` | `text-generation` |
| {doc}`olmo3` | `OLMo2CausalLMModel` | `text-generation` |
| {doc}`open-llama` | `CausalLMModel` | `text-generation` |
| {doc}`openelm` | `CausalLMModel` | `text-generation` |
| {doc}`persimmon` | `PersimmonCausalLMModel` | `text-generation` |
| {doc}`phi` | `PhiCausalLMModel` | `text-generation` |
| {doc}`phi3` | `Phi3CausalLMModel` | `text-generation` |
| {doc}`phi3small` | `Phi3SmallCausalLMModel` | `text-generation` |
| {doc}`qwen` | `QwenCausalLMModel` | `text-generation` |
| {doc}`qwen2` | `CausalLMModel` | `text-generation` |
| {doc}`qwen2_5_vl_text` | `Qwen25VLTextModel` | `text-generation` |
| {doc}`qwen2_vl_text` | `Qwen25VLTextModel` | `text-generation` |
| {doc}`qwen3` | `Qwen3CausalLMModel` | `text-generation` |
| {doc}`qwen3_5_text` | `Qwen35CausalLMModel` | `hybrid-text-generation` |
| {doc}`qwen3_5_vl_text` | `Qwen35VLTextModel` | `hybrid-text-generation` |
| {doc}`qwen3_tts_tokenizer_12hz` | `Qwen3TTSTokenizerV2Model` | `text-generation` |
| {doc}`qwen3_vl_text` | `Qwen3VLTextModel` | `text-generation` |
| {doc}`seed_oss` | `CausalLMModel` | `text-generation` |
| {doc}`shieldgemma2` | `Gemma2CausalLMModel` | `text-generation` |
| {doc}`smollm3` | `SmolLM3CausalLMModel` | `text-generation` |
| {doc}`solar_open` | `CausalLMModel` | `text-generation` |
| {doc}`stablelm` | `LayerNormCausalLMModel` | `text-generation` |
| {doc}`starcoder2` | `StarCoder2CausalLMModel` | `text-generation` |
| {doc}`yi` | `CausalLMModel` | `text-generation` |
| {doc}`zamba` | `CausalLMModel` | `text-generation` |

## Mixture of Experts

Models that route tokens to a subset of expert MLPs.

| `model_type` | Class | Task |
|---|---|---|
| {doc}`arctic` | `MoECausalLMModel` | `text-generation` |
| {doc}`dbrx` | `MoECausalLMModel` | `text-generation` |
| {doc}`deepseek_v2` | `DeepSeekV3CausalLMModel` | `text-generation` |
| {doc}`deepseek_v2_moe` | `DeepSeekV3CausalLMModel` | `text-generation` |
| {doc}`deepseek_v3` | `DeepSeekV3CausalLMModel` | `text-generation` |
| {doc}`deepseek_v4` | `DeepSeekV4CausalLMModel` | `deepseek-v4` |
| {doc}`dots1` | `DeepSeekV3CausalLMModel` | `text-generation` |
| {doc}`ernie4_5_moe` | `Ernie45MoECausalLMModel` | `text-generation` |
| {doc}`flex_olmo` | `MoECausalLMModel` | `text-generation` |
| {doc}`glm4_moe` | `Glm4MoECausalLMModel` | `text-generation` |
| {doc}`glm4v_moe_text` | `Glm4MoECausalLMModel` | `text-generation` |
| {doc}`gpt_oss` | `GPTOSSCausalLMModel` | `text-generation` |
| {doc}`granitemoe` | `GraniteMoECausalLMModel` | `text-generation` |
| {doc}`granitemoeshared` | `GraniteMoECausalLMModel` | `text-generation` |
| {doc}`hunyuan_v1_moe` | `HunYuanMoEV1CausalLMModel` | `text-generation` |
| {doc}`jetmoe` | `JetMoeCausalLMModel` | `text-generation` |
| {doc}`longcat_flash` | `LongcatFlashCausalLMModel` | `text-generation` |
| {doc}`mixtral` | `MoECausalLMModel` | `text-generation` |
| {doc}`olmoe` | `MoECausalLMModel` | `text-generation` |
| {doc}`phimoe` | `Phi3MoECausalLMModel` | `text-generation` |
| {doc}`qwen2_moe` | `Qwen2MoECausalLMModel` | `text-generation` |
| {doc}`qwen3_5_moe` | `Qwen35MoECausalLMModel` | `hybrid-text-generation` |
| {doc}`qwen3_moe` | `MoECausalLMModel` | `text-generation` |
| {doc}`qwen3_next` | `Qwen3NextCausalLMModel` | `hybrid-text-generation` |
| {doc}`qwen3_omni_moe` | `MoECausalLMModel` | `text-generation` |
| {doc}`qwen3_vl_moe` | `MoECausalLMModel` | `text-generation` |
| {doc}`youtu` | `DeepSeekV3CausalLMModel` | `text-generation` |

## Multimodal

Models that process images, audio, or other modalities alongside text.

| `model_type` | Class | Task |
|---|---|---|
| {doc}`aya_vision` | `LLaVAModel` | `vision-language` |
| {doc}`blip-2` | `Blip2Model` | `vision-language` |
| {doc}`chameleon` | `LLaVAModel` | `vision-language` |
| {doc}`cohere2_vision` | `LLaVAModel` | `vision-language` |
| {doc}`deepseek_vl` | `LLaVAModel` | `vision-language` |
| {doc}`deepseek_vl_hybrid` | `LLaVAModel` | `vision-language` |
| {doc}`deepseek_vl_v2` | `DeepSeekOCR2CausalLMModel` | `vision-language` |
| {doc}`florence2` | `LLaVAModel` | `vision-language` |
| {doc}`fuyu` | `LLaVAModel` | `vision-language` |
| {doc}`gemma3` | `Gemma3MultiModalModel` | `vision-language` |
| {doc}`gemma4` | `Gemma4Model` | `gemma4` |
| {doc}`gemma4_unified` | `Gemma4UnifiedModel` | `gemma4-unified` |
| {doc}`glm4v` | `LLaVAModel` | `vision-language` |
| {doc}`glm4v_moe` | `LLaVAModel` | `vision-language` |
| {doc}`got_ocr2` | `LLaVAModel` | `vision-language` |
| {doc}`hunyuan_vl_mot` | `HunYuanVLMoTModel` | `hunyuan-vl-mot` |
| {doc}`idefics2` | `LLaVAModel` | `vision-language` |
| {doc}`idefics3` | `LLaVAModel` | `vision-language` |
| {doc}`instructblip` | `LLaVAModel` | `vision-language` |
| {doc}`instructblipvideo` | `LLaVAModel` | `vision-language` |
| {doc}`internvl` | `InternVL2Model` | `vision-language` |
| {doc}`internvl2` | `InternVL2Model` | `vision-language` |
| {doc}`internvl_chat` | `InternVL2Model` | `vision-language` |
| {doc}`janus` | `LLaVAModel` | `vision-language` |
| {doc}`llava` | `LLaVAModel` | `vision-language` |
| {doc}`llava_next` | `LLaVAModel` | `vision-language` |
| {doc}`llava_next_video` | `LLaVAModel` | `vision-language` |
| {doc}`llava_onevision` | `LLaVAModel` | `vision-language` |
| {doc}`mistral3` | `LLaVAModel` | `vision-language` |
| {doc}`mllama` | `MllamaCausalLMModel` | `mllama-vision-language` |
| {doc}`molmo` | `LLaVAModel` | `vision-language` |
| {doc}`ovis2` | `LLaVAModel` | `vision-language` |
| {doc}`paligemma` | `LLaVAModel` | `vision-language` |
| {doc}`phi3_v` | `Phi3VModel` | `vision-language` |
| {doc}`phi4-siglip` | `Phi4SigLIPModel` | `vision-language` |
| {doc}`phi4_multimodal` | `Phi4MMMultiModalModel` | `phi4mm-multimodal` |
| {doc}`phi4mm` | `Phi4MMMultiModalModel` | `phi4mm-multimodal` |
| {doc}`pixtral` | `LLaVAModel` | `vision-language` |
| {doc}`qwen2_5_vl` | `Qwen25VLCausalLMModel` | `qwen-vl` |
| {doc}`qwen2_vl` | `Qwen2VLCausalLMModel` | `qwen-vl` |
| {doc}`qwen3_5` | `Qwen35VL3ModelCausalLMModel` | `hybrid-qwen-vl` |
| {doc}`qwen3_5_moe_vl` | `Qwen35MoEVL3ModelCausalLMModel` | `hybrid-qwen-vl` |
| {doc}`qwen3_5_vl` | `Qwen35VL3ModelCausalLMModel` | `hybrid-qwen-vl` |
| {doc}`qwen3_vl` | `Qwen3VL3ModelCausalLMModel` | `qwen-vl` |
| {doc}`qwen3_vl_single` | `Qwen3VLCausalLMModel` | `qwen3-vl-vision-language` |
| {doc}`smolvlm` | `LLaVAModel` | `vision-language` |
| {doc}`video_llava` | `LLaVAModel` | `vision-language` |
| {doc}`vipllava` | `LLaVAModel` | `vision-language` |

## Speech-to-Text

Encoder-decoder models for speech recognition.

| `model_type` | Class | Task |
|---|---|---|
| {doc}`fastconformer_rnnt` | `EncDecRNNTModel` | `fastconformer-rnnt` |
| {doc}`fun_asr` | `FunASRForConditionalGeneration` | `fun-asr-speech-language` |
| {doc}`mms` | `Wav2Vec2ForCTCModel` | `ctc-asr` |
| {doc}`qwen3_asr` | `Qwen3ASRForConditionalGeneration` | `speech-language` |
| {doc}`qwen3_forced_aligner` | `Qwen3ASRForConditionalGeneration` | `speech-language` |
| {doc}`sensevoice_small` | `SenseVoiceSmallModel` | `audio-ctc` |
| {doc}`whisper` | `WhisperForConditionalGeneration` | `speech-to-text` |

## Audio

Audio encoder models for feature extraction (Wav2Vec2, HuBERT, WavLM).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`data2vec-audio` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`hubert` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`mctct` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`musicgen` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`qwen3_tts` | `Qwen3TTSForConditionalGeneration` | `tts` |
| {doc}`seamless_m4t` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`seamless_m4t_v2` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`sew` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`sew-d` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`speecht5` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`unispeech` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`unispeech-sat` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`voxtral_encoder` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`wav2vec2` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`wav2vec2-bert` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`wav2vec2-conformer` | `Wav2Vec2Model` | `audio-feature-extraction` |
| {doc}`wavlm` | `Wav2Vec2Model` | `audio-feature-extraction` |

## encoder-only

Encoder-only models for embeddings and classification (BERT, RoBERTa).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`distilbert` | `DistilBertModel` | `feature-extraction` |

## encoder

Encoder-only transformer models (BERT family).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`albert` | `BertModel` | `feature-extraction` |
| {doc}`bert` | `BertModel` | `feature-extraction` |
| {doc}`bros` | `BertModel` | `feature-extraction` |
| {doc}`camembert` | `BertModel` | `feature-extraction` |
| {doc}`data2vec-text` | `BertModel` | `feature-extraction` |
| {doc}`deberta` | `BertModel` | `feature-extraction` |
| {doc}`deberta-v2` | `BertModel` | `feature-extraction` |
| {doc}`electra` | `BertModel` | `feature-extraction` |
| {doc}`ernie` | `BertModel` | `feature-extraction` |
| {doc}`ernie_m` | `BertModel` | `feature-extraction` |
| {doc}`esm` | `BertModel` | `feature-extraction` |
| {doc}`flaubert` | `BertModel` | `feature-extraction` |
| {doc}`ibert` | `BertModel` | `feature-extraction` |
| {doc}`layoutlm` | `BertModel` | `feature-extraction` |
| {doc}`layoutlmv2` | `BertModel` | `feature-extraction` |
| {doc}`layoutlmv3` | `LayoutLMv3Model` | `feature-extraction` |
| {doc}`lilt` | `BertModel` | `feature-extraction` |
| {doc}`markuplm` | `BertModel` | `feature-extraction` |
| {doc}`mega` | `BertModel` | `feature-extraction` |
| {doc}`megatron-bert` | `BertModel` | `feature-extraction` |
| {doc}`mobilebert` | `BertModel` | `feature-extraction` |
| {doc}`modernbert` | `ModernBertModel` | `feature-extraction` |
| {doc}`mpnet` | `BertModel` | `feature-extraction` |
| {doc}`mra` | `BertModel` | `feature-extraction` |
| {doc}`nezha` | `BertModel` | `feature-extraction` |
| {doc}`nystromformer` | `BertModel` | `feature-extraction` |
| {doc}`qdqbert` | `BertModel` | `feature-extraction` |
| {doc}`rembert` | `BertModel` | `feature-extraction` |
| {doc}`roberta` | `BertModel` | `feature-extraction` |
| {doc}`roberta-prelayernorm` | `BertModel` | `feature-extraction` |
| {doc}`roc_bert` | `BertModel` | `feature-extraction` |
| {doc}`roformer` | `BertModel` | `feature-extraction` |
| {doc}`splinter` | `BertModel` | `feature-extraction` |
| {doc}`squeezebert` | `BertModel` | `feature-extraction` |
| {doc}`xlm-roberta` | `BertModel` | `feature-extraction` |
| {doc}`xlm-roberta-xl` | `BertModel` | `feature-extraction` |
| {doc}`xlnet` | `BertModel` | `feature-extraction` |
| {doc}`xmod` | `BertModel` | `feature-extraction` |
| {doc}`yoso` | `BertModel` | `feature-extraction` |

## encoder-decoder

Encoder-decoder sequence-to-sequence models (BART, T5, mBART).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`bart` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`bigbird_pegasus` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`blenderbot` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`blenderbot-small` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`fsmt` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`led` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`longt5` | `T5ForConditionalGeneration` | `seq2seq` |
| {doc}`m2m_100` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`marian` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`mbart` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`mt5` | `T5ForConditionalGeneration` | `seq2seq` |
| {doc}`mvp` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`nllb-moe` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`nllb_moe` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`pegasus` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`pegasus_x` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`plbart` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`prophetnet` | `BartForConditionalGeneration` | `seq2seq` |
| {doc}`switch_transformers` | `T5ForConditionalGeneration` | `seq2seq` |
| {doc}`t5` | `T5ForConditionalGeneration` | `seq2seq` |
| {doc}`trocr` | `TrOCRForConditionalGeneration` | `seq2seq` |
| {doc}`umt5` | `T5ForConditionalGeneration` | `seq2seq` |
| {doc}`xlm-prophetnet` | `BartForConditionalGeneration` | `seq2seq` |

## causal-lm

Absolute positional embedding language models (GPT-2 style).

| `model_type` | Class | Task |
|---|---|---|
| {doc}`biogpt` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`ctrl` | `CTRLCausalLMModel` | `text-generation` |
| {doc}`gpt-sw3` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`gpt2` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`gpt_bigcode` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`gpt_neo` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`imagegpt` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`openai-gpt` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`opt` | `OPTCausalLMModel` | `text-generation` |
| {doc}`xglm` | `GPT2CausalLMModel` | `text-generation` |
| {doc}`xlm` | `XLMCausalLMModel` | `text-generation` |

## vision

Vision-only models for image classification and feature extraction.

| `model_type` | Class | Task |
|---|---|---|
| {doc}`beit` | `ViTModel` | `image-classification` |
| {doc}`blip` | `BlipVisionModel` | `image-classification` |
| {doc}`clip_vision_model` | `CLIPVisionModel` | `image-classification` |
| {doc}`cvt` | `ViTModel` | `image-classification` |
| {doc}`data2vec-vision` | `ViTModel` | `image-classification` |
| {doc}`deit` | `ViTModel` | `image-classification` |
| {doc}`dinov2` | `ViTModel` | `image-classification` |
| {doc}`dinov2_with_registers` | `ViTModel` | `image-classification` |
| {doc}`dinov3_vit` | `ViTModel` | `image-classification` |
| {doc}`hiera` | `ViTModel` | `image-classification` |
| {doc}`ijepa` | `ViTModel` | `image-classification` |
| {doc}`mobilevit` | `ViTModel` | `image-classification` |
| {doc}`mobilevitv2` | `ViTModel` | `image-classification` |
| {doc}`pvt` | `ViTModel` | `image-classification` |
| {doc}`pvt_v2` | `ViTModel` | `image-classification` |
| {doc}`sam2` | `Sam2VisionModel` | `image-classification` |
| {doc}`siglip` | `SigLIPVisionModel` | `image-classification` |
| {doc}`siglip2` | `SigLIPVisionModel` | `image-classification` |
| {doc}`siglip2_vision_model` | `SigLIPVisionModel` | `image-classification` |
| {doc}`siglip_vision_model` | `SigLIPVisionModel` | `image-classification` |
| {doc}`swin` | `ViTModel` | `image-classification` |
| {doc}`swin2sr` | `ViTModel` | `image-classification` |
| {doc}`swinv2` | `ViTModel` | `image-classification` |
| {doc}`vit` | `ViTModel` | `image-classification` |
| {doc}`vit_hybrid` | `ViTModel` | `image-classification` |
| {doc}`vit_mae` | `ViTModel` | `image-classification` |
| {doc}`vit_msn` | `ViTModel` | `image-classification` |

## Depth Estimation

| `model_type` | Class | Task |
|---|---|---|
| {doc}`depth_anything` | `DepthAnythingForDepthEstimation` | `image-classification` |

## Encoder

| `model_type` | Class | Task |
|---|---|---|
| {doc}`clip_text_model` | `CLIPTextModel` | `feature-extraction` |

## Hybrid SSM+Attention

| `model_type` | Class | Task |
|---|---|---|
| {doc}`bamba` | `BambaCausalLMModel` | `hybrid-text-generation` |
| {doc}`granitemoehybrid` | `GraniteMoeHybridCausalLMModel` | `hybrid-text-generation` |
| {doc}`jamba` | `JambaCausalLMModel` | `hybrid-text-generation` |
| {doc}`nemotron_h` | `NemotronHCausalLMModel` | `hybrid-text-generation` |
| {doc}`zamba2` | `Zamba2CausalLMModel` | `hybrid-text-generation` |

## Object Detection

| `model_type` | Class | Task |
|---|---|---|
| {doc}`yolos` | `YolosForObjectDetection` | `object-detection` |

## Segmentation

| `model_type` | Class | Task |
|---|---|---|
| {doc}`segformer` | `SegformerForSemanticSegmentation` | `image-classification` |

## Text

| `model_type` | Class | Task |
|---|---|---|
| {doc}`gemma4_text` | `Gemma4CausalLMModel` | `gemma4-text-generation` |
| {doc}`gemma4_unified_text` | `Gemma4CausalLMModel` | `gemma4-text-generation` |

```{toctree}
:maxdepth: 1
:hidden:

DFlashDraftModel
Eagle3DraftModel
Eagle3LlamaForCausalLM
Eagle3Speculator
Gemma4AssistantForCausalLM
Gemma4UnifiedAssistantForCausalLM
LlamaForCausalLMEagle3
Qwen35MtpModel
albert
apertus
arcee
arctic
aya_vision
baichuan
bamba
bart
beit
bert
bigbird_pegasus
biogpt
blenderbot
blenderbot-small
blip
blip-2
bloom
bros
camembert
chameleon
chatglm
clip_text_model
clip_vision_model
code_llama
codegen
codegen2
cohere
cohere2
cohere2_vision
command_r
csm
ctrl
cvt
data2vec-audio
data2vec-text
data2vec-vision
dbrx
deberta
deberta-v2
deepseek_v2
deepseek_v2_moe
deepseek_v3
deepseek_v4
deepseek_vl
deepseek_vl_hybrid
deepseek_vl_v2
deit
depth_anything
diffllama
dinov2
dinov2_with_registers
dinov3_vit
distilbert
doge
dots1
electra
ernie
ernie4_5
ernie4_5_moe
ernie_m
esm
evolla
exaone
exaone4
falcon
falcon_h1
falcon_mamba
fastconformer_rnnt
flaubert
flex_olmo
florence2
fsmt
fun_asr
fuyu
gemma
gemma2
gemma3
gemma3_text
gemma3n
gemma3n_text
gemma4
gemma4_assistant
gemma4_text
gemma4_unified
gemma4_unified_assistant
gemma4_unified_text
glm
glm4
glm4_moe
glm4v
glm4v_moe
glm4v_moe_text
glm4v_text
got_ocr2
gpt-sw3
gpt2
gpt_bigcode
gpt_neo
gpt_neox
gpt_neox_japanese
gpt_oss
gptj
granite
granitemoe
granitemoehybrid
granitemoeshared
helium
hiera
hubert
hunyuan_v1_dense
hunyuan_v1_moe
hunyuan_vl_mot
ibert
idefics2
idefics3
ijepa
imagegpt
instructblip
instructblipvideo
internlm2
internvl
internvl2
internvl_chat
jamba
janus
jetmoe
layoutlm
layoutlmv2
layoutlmv3
led
lilt
llada
llama
llama4_text
llava
llava_next
llava_next_video
llava_onevision
longcat_flash
longt5
m2m_100
mamba
mamba2
marian
markuplm
mbart
mctct
mega
megatron-bert
minicpm
minicpm3
minimax
ministral
ministral3
mistral
mistral3
mixtral
mllama
mms
mobilebert
mobilevit
mobilevitv2
modernbert
modernbert-decoder
molmo
mpnet
mpt
mra
mt5
musicgen
mvp
nanochat
nemotron
nemotron_h
nezha
nllb-moe
nllb_moe
nystromformer
olmo
olmo2
olmo3
olmoe
open-llama
openai-gpt
openelm
opt
ovis2
paligemma
pegasus
pegasus_x
persimmon
phi
phi3
phi3_v
phi3small
phi4-siglip
phi4_multimodal
phi4mm
phimoe
pixtral
plbart
prophetnet
pvt
pvt_v2
qdqbert
qwen
qwen2
qwen2_5_vl
qwen2_5_vl_text
qwen2_moe
qwen2_vl
qwen2_vl_text
qwen3
qwen3_5
qwen3_5_moe
qwen3_5_moe_vl
qwen3_5_text
qwen3_5_vl
qwen3_5_vl_text
qwen3_asr
qwen3_forced_aligner
qwen3_moe
qwen3_next
qwen3_omni_moe
qwen3_tts
qwen3_tts_tokenizer_12hz
qwen3_vl
qwen3_vl_moe
qwen3_vl_single
qwen3_vl_text
rembert
roberta
roberta-prelayernorm
roc_bert
roformer
sam2
seamless_m4t
seamless_m4t_v2
seed_oss
segformer
sensevoice_small
sew
sew-d
shieldgemma2
siglip
siglip2
siglip2_vision_model
siglip_vision_model
smollm3
smolvlm
solar_open
speecht5
splinter
squeezebert
stablelm
starcoder2
swin
swin2sr
swinv2
switch_transformers
t5
trocr
umt5
unispeech
unispeech-sat
video_llava
vipllava
vit
vit_hybrid
vit_mae
vit_msn
voxtral_encoder
wav2vec2
wav2vec2-bert
wav2vec2-conformer
wavlm
whisper
xglm
xlm
xlm-prophetnet
xlm-roberta
xlm-roberta-xl
xlnet
xmod
yi
yolos
yoso
youtu
zamba
zamba2
```