Supported Models — vLLM (original) (raw)

Supported Models#

vLLM supports generative and pooling models across various tasks. If a model supports more than one task, you can set the task via the --task argument.

For each task, we list the model architectures that have been implemented in vLLM. Alongside each architecture, we include some popular models that use it.

Model Implementation#

vLLM#

If vLLM natively supports a model, its implementation can be found in vllm/model_executor/models.

These models are what we list in List of Text-only Language Models and List of Multimodal Language Models.

Transformers#

vLLM also supports model implementations that are available in Transformers. This does not currently work for all models, but most decoder language models are supported, and vision language model support is planned!

To check if the modeling backend is Transformers, you can simply do this:

from vllm import LLM llm = LLM(model=..., task="generate") # Name or path of your model llm.apply_model(lambda model: print(type(model)))

If it is TransformersForCausalLM then it means it’s based on Transformers!

Note

vLLM may not fully optimise the Transformers implementation so you may see degraded performance if comparing a native model to a Transformers model in vLLM.

Custom models#

If a model is neither supported natively by vLLM or Transformers, it can still be used in vLLM!

For a model to be compatible with the Transformers backend for vLLM it must:

• be a Transformers compatible custom model (see Transformers - Customizing models):
  • The model directory must have the correct structure (e.g. config.json is present).
  • config.json must contain auto_map.AutoModel.
• be a Transformers backend for vLLM compatible model (see Writing custom models):
  • Customisation should be done in the base model (e.g. in MyModel, not MyModelForCausalLM).

If the compatible model is:

• on the Hugging Face Model Hub, simply set trust_remote_code=True for Offline Inference or --trust-remode-code for the OpenAI-Compatible Server.
• in a local directory, simply pass directory path to model=<MODEL_DIR> for Offline Inference or vllm serve <MODEL_DIR> for the OpenAI-Compatible Server.

This means that, with the Transformers backend for vLLM, new models can be used before they are officially supported in Transformers or vLLM!

Writing custom models#

This section details the necessary modifications to make to a Transformers compatible custom model that make it compatible with the Transformers backend for vLLM. (We assume that a Transformers compatible custom model has already been created, see Transformers - Customizing models).

To make your model compatible with the Transformers backend, it needs:

1. kwargs passed down through all modules from MyModel to MyAttention.
2. MyAttention must use ALL_ATTENTION_FUNCTIONS to call attention.
3. MyModel must contain _supports_attention_backend = True.

modeling_my_model.py#

from transformers import PreTrainedModel from torch import nn

class MyAttention(nn.Module):

def forward(self, hidden_states, **kwargs): ... attention_interface = ALL_ATTENTION_FUNCTIONS[self.config._attn_implementation] attn_output, attn_weights = attention_interface( self, query_states, key_states, value_states, **kwargs, ) ...

class MyModel(PreTrainedModel): _supports_attention_backend = True

Here is what happens in the background when this model is loaded:

1. The config is loaded.
2. MyModel Python class is loaded from the auto_map in config, and we check that the model is_backend_compatible().
3. MyModel is loaded into TransformersForCausalLM (see vllm/model_executor/models/transformers.py) which sets self.config._attn_implementation = "vllm" so that vLLM’s attention layer is used.

That’s it!

For your model to be compatible with vLLM’s tensor parallel and/or pipeline parallel features, you must add base_model_tp_plan and/or base_model_pp_plan to your model’s config class:

configuration_my_model.py#

from transformers import PretrainedConfig

class MyConfig(PretrainedConfig): base_model_tp_plan = { "layers..self_attn.k_proj": "colwise", "layers..self_attn.v_proj": "colwise", "layers..self_attn.o_proj": "rowwise", "layers..mlp.gate_proj": "colwise", "layers..mlp.up_proj": "colwise", "layers..mlp.down_proj": "rowwise", } base_model_pp_plan = { "embed_tokens": (["input_ids"], ["inputs_embeds"]), "layers": (["hidden_states", "attention_mask"], ["hidden_states"]), "norm": (["hidden_states"], ["hidden_states"]), }

• base_model_tp_plan is a dict that maps fully qualified layer name patterns to tensor parallel styles (currently only "colwise" and "rowwise" are supported).
• base_model_pp_plan is a dict that maps direct child layer names to tuples of lists of strs:
  • You only need to do this for layers which are not present on all pipeline stages
  • vLLM assumes that there will be only one nn.ModuleList, which is distributed across the pipeline stages
  • The list in the first element of the tuple contains the names of the input arguments
  • The list in the last element of the tuple contains the names of the variables the layer outputs to in your modeling code

Loading a Model#

Hugging Face Hub#

By default, vLLM loads models from Hugging Face (HF) Hub. To change the download path for models, you can set the HF_HOME environment variable; for more details, refer to their official documentation.

To determine whether a given model is natively supported, you can check the config.json file inside the HF repository. If the "architectures" field contains a model architecture listed below, then it should be natively supported.

Models do not need to be natively supported to be used in vLLM. The Transformers backend enables you to run models directly using their Transformers implementation (or even remote code on the Hugging Face Model Hub!).

Tip

The easiest way to check if your model is really supported at runtime is to run the program below:

from vllm import LLM

For generative models (task=generate) only

llm = LLM(model=..., task="generate") # Name or path of your model output = llm.generate("Hello, my name is") print(output)

For pooling models (task={embed,classify,reward,score}) only

llm = LLM(model=..., task="embed") # Name or path of your model output = llm.encode("Hello, my name is") print(output)

If vLLM successfully returns text (for generative models) or hidden states (for pooling models), it indicates that your model is supported.

Otherwise, please refer to Adding a New Model for instructions on how to implement your model in vLLM. Alternatively, you can open an issue on GitHub to request vLLM support.

Using a proxy#

Here are some tips for loading/downloading models from Hugging Face using a proxy:

• Set the proxy globally for your session (or set it in the profile file):

export http_proxy=http://your.proxy.server:port export https_proxy=http://your.proxy.server:port

• Set the proxy for just the current command:

https_proxy=http://your.proxy.server:port huggingface-cli download

or use vllm cmd directly

https_proxy=http://your.proxy.server:port vllm serve --disable-log-requests

• Set the proxy in Python interpreter:

import os

os.environ['http_proxy'] = 'http://your.proxy.server:port' os.environ['https_proxy'] = 'http://your.proxy.server:port'

ModelScope#

To use models from ModelScope instead of Hugging Face Hub, set an environment variable:

export VLLM_USE_MODELSCOPE=True

And use with trust_remote_code=True.

from vllm import LLM

llm = LLM(model=..., revision=..., task=..., trust_remote_code=True)

For generative models (task=generate) only

output = llm.generate("Hello, my name is") print(output)

For pooling models (task={embed,classify,reward,score}) only

output = llm.encode("Hello, my name is") print(output)

Feature Status Legend#

• ✅︎ indicates that the feature is supported for the model.
• 🚧 indicates that the feature is planned but not yet supported for the model.
• ⚠️ indicates that the feature is available but may have known issues or limitations.

List of Text-only Language Models#

Generative Models#

See this page for more information on how to use generative models.

Text Generation (--task generate)#

Note

Currently, the ROCm version of vLLM supports Mistral and Mixtral only for context lengths up to 4096.

Pooling Models#

See this page for more information on how to use pooling models.

Important

Since some model architectures support both generative and pooling tasks, you should explicitly specify the task type to ensure that the model is used in pooling mode instead of generative mode.

Text Embedding (--task embed)#

Note

ssmits/Qwen2-7B-Instruct-embed-base has an improperly defined Sentence Transformers config. You should manually set mean pooling by passing --override-pooler-config '{"pooling_type": "MEAN"}'.

Note

The HF implementation of Alibaba-NLP/gte-Qwen2-1.5B-instruct is hardcoded to use causal attention despite what is shown in config.json. To compare vLLM vs HF results, you should set --hf-overrides '{"is_causal": true}' in vLLM so that the two implementations are consistent with each other.

For both the 1.5B and 7B variants, you also need to enable --trust-remote-code for the correct tokenizer to be loaded. See relevant issue on HF Transformers.

If your model is not in the above list, we will try to automatically convert the model usingas_embedding_model(). By default, the embeddings of the whole prompt are extracted from the normalized hidden state corresponding to the last token.

Reward Modeling (--task reward)#

If your model is not in the above list, we will try to automatically convert the model usingas_reward_model(). By default, we return the hidden states of each token directly.

Important

For process-supervised reward models such as peiyi9979/math-shepherd-mistral-7b-prm, the pooling config should be set explicitly, e.g.: --override-pooler-config '{"pooling_type": "STEP", "step_tag_id": 123, "returned_token_ids": [456, 789]}'.

Classification (--task classify)#

If your model is not in the above list, we will try to automatically convert the model usingas_classification_model(). By default, the class probabilities are extracted from the softmaxed hidden state corresponding to the last token.

Sentence Pair Scoring (--task score)#

List of Multimodal Language Models#

The following modalities are supported depending on the model:

• Text
• Image
• Video
• Audio

Any combination of modalities joined by + are supported.

• e.g.: T + I means that the model supports text-only, image-only, and text-with-image inputs.

On the other hand, modalities separated by / are mutually exclusive.

• e.g.: T / I means that the model supports text-only and image-only inputs, but not text-with-image inputs.

See this page on how to pass multi-modal inputs to the model.

Important

To enable multiple multi-modal items per text prompt in vLLM V0, you have to set limit_mm_per_prompt (offline inference) or --limit-mm-per-prompt (online serving). For example, to enable passing up to 4 images per text prompt:

Offline inference:

from vllm import LLM

llm = LLM( model="Qwen/Qwen2-VL-7B-Instruct", limit_mm_per_prompt={"image": 4}, )

Online serving:

vllm serve Qwen/Qwen2-VL-7B-Instruct --limit-mm-per-prompt '{"image":4}'

This is no longer required if you are using vLLM V1.

Note

vLLM currently only supports adding LoRA to the language backbone of multimodal models.

Generative Models#

See this page for more information on how to use generative models.

Text Generation (--task generate)#

^ You need to set the architecture name via --hf-overrides to match the one in vLLM.
• For example, to use DeepSeek-VL2 series models:
--hf-overrides '{"architectures": ["DeepseekVLV2ForCausalLM"]}'
E Pre-computed embeddings can be inputted for this modality.

• Multiple items can be inputted per text prompt for this modality.

Important

Pan-and-scan image pre-processing is currently supported on V0 (but not V1). You can enable it by passing --mm-processor-kwargs '{"do_pan_and_scan": true}'.

Warning

Both V0 and V1 support Gemma3ForConditionalGeneration for text-only inputs. However, there are differences in how they handle text + image inputs:

V0 correctly implements the model’s attention pattern:

• Uses bidirectional attention between the image tokens corresponding to the same image
• Uses causal attention for other tokens
• Implemented via (naive) PyTorch SDPA with masking tensors
• Note: May use significant memory for long prompts with image

V1 currently uses a simplified attention pattern:

• Uses causal attention for all tokens, including image tokens
• Generates reasonable outputs but does not match the original model’s attention for text + image inputs, especially when {"do_pan_and_scan": true}
• Will be updated in the future to support the correct behavior

This limitation exists because the model’s mixed attention pattern (bidirectional for images, causal otherwise) is not yet supported by vLLM’s attention backends.

Note

h2oai/h2ovl-mississippi-2b will be available in V1 once we support backends other than FlashAttention.

Note

To use TIGER-Lab/Mantis-8B-siglip-llama3, you have to pass --hf_overrides '{"architectures": ["MantisForConditionalGeneration"]}' when running vLLM.

Warning

The output quality of AllenAI/Molmo-7B-D-0924 (especially in object localization tasks) has deteriorated in recent updates.

For the best results, we recommend using the following dependency versions (tested on A10 and L40):

Core vLLM-compatible dependencies with Molmo accuracy setup (tested on L40)

torch==2.5.1 torchvision==0.20.1 transformers==4.48.1 tokenizers==0.21.0 tiktoken==0.7.0 vllm==0.7.0

Optional but recommended for improved performance and stability

triton==3.1.0 xformers==0.0.28.post3 uvloop==0.21.0 protobuf==5.29.3 openai==1.60.2 opencv-python-headless==4.11.0.86 pillow==10.4.0

Installed FlashAttention (for float16 only)

flash-attn>=2.5.6 # Not used in float32, but should be documented

Note: Make sure you understand the security implications of using outdated packages.

Note

The official openbmb/MiniCPM-V-2 doesn’t work yet, so we need to use a fork (HwwwH/MiniCPM-V-2) for now. For more details, please see: Pull Request #4087

Warning

Our PaliGemma implementations have the same problem as Gemma 3 (see above) for both V0 and V1.

Note

To use Qwen2.5-Omni, you have to install Hugging Face Transformers library from source viapip install git+https://github.com/huggingface/transformers.git.

Read audio from video pre-processing is currently supported on V0 (but not V1), because overlapping modalities is not yet supported in V1.--mm-processor-kwargs '{"use_audio_in_video": true}'.

Pooling Models#

See this page for more information on how to use pooling models.

Important

Since some model architectures support both generative and pooling tasks, you should explicitly specify the task type to ensure that the model is used in pooling mode instead of generative mode.

Text Embedding (--task embed)#

Any text generation model can be converted into an embedding model by passing --task embed.

Note

To get the best results, you should use pooling models that are specifically trained as such.

The following table lists those that are tested in vLLM.

Transcription (--task transcription)#

Speech2Text models trained specifically for Automatic Speech Recognition.

Model Support Policy#

At vLLM, we are committed to facilitating the integration and support of third-party models within our ecosystem. Our approach is designed to balance the need for robustness and the practical limitations of supporting a wide range of models. Here’s how we manage third-party model support:

1. Community-Driven Support: We encourage community contributions for adding new models. When a user requests support for a new model, we welcome pull requests (PRs) from the community. These contributions are evaluated primarily on the sensibility of the output they generate, rather than strict consistency with existing implementations such as those in transformers. Call for contribution: PRs coming directly from model vendors are greatly appreciated!
2. Best-Effort Consistency: While we aim to maintain a level of consistency between the models implemented in vLLM and other frameworks like transformers, complete alignment is not always feasible. Factors like acceleration techniques and the use of low-precision computations can introduce discrepancies. Our commitment is to ensure that the implemented models are functional and produce sensible results.
   Tip
   When comparing the output of model.generate from Hugging Face Transformers with the output of llm.generate from vLLM, note that the former reads the model’s generation config file (i.e., generation_config.json) and applies the default parameters for generation, while the latter only uses the parameters passed to the function. Ensure all sampling parameters are identical when comparing outputs.
3. Issue Resolution and Model Updates: Users are encouraged to report any bugs or issues they encounter with third-party models. Proposed fixes should be submitted via PRs, with a clear explanation of the problem and the rationale behind the proposed solution. If a fix for one model impacts another, we rely on the community to highlight and address these cross-model dependencies. Note: for bugfix PRs, it is good etiquette to inform the original author to seek their feedback.
4. Monitoring and Updates: Users interested in specific models should monitor the commit history for those models (e.g., by tracking changes in the main/vllm/model_executor/models directory). This proactive approach helps users stay informed about updates and changes that may affect the models they use.
5. Selective Focus: Our resources are primarily directed towards models with significant user interest and impact. Models that are less frequently used may receive less attention, and we rely on the community to play a more active role in their upkeep and improvement.

Through this approach, vLLM fosters a collaborative environment where both the core development team and the broader community contribute to the robustness and diversity of the third-party models supported in our ecosystem.

Note that, as an inference engine, vLLM does not introduce new models. Therefore, all models supported by vLLM are third-party models in this regard.

We have the following levels of testing for models:

1. Strict Consistency: We compare the output of the model with the output of the model in the HuggingFace Transformers library under greedy decoding. This is the most stringent test. Please refer to models tests for the models that have passed this test.
2. Output Sensibility: We check if the output of the model is sensible and coherent, by measuring the perplexity of the output and checking for any obvious errors. This is a less stringent test.
3. Runtime Functionality: We check if the model can be loaded and run without errors. This is the least stringent test. Please refer to functionality tests and examples for the models that have passed this test.
4. Community Feedback: We rely on the community to provide feedback on the models. If a model is broken or not working as expected, we encourage users to raise issues to report it or open pull requests to fix it. The rest of the models fall under this category.