# Offline Inference Offline inference examples demonstrate how to use vLLM in an offline setting, where the model is queried for predictions in batches. We recommend starting with . :::{toctree} :caption: Examples :maxdepth: 1 audio_language basic chat_with_tools cpu_offload_lmcache data_parallel disaggregated_prefill disaggregated_prefill_lmcache distributed eagle embed_jina_embeddings_v3 embed_matryoshka_fy encoder_decoder encoder_decoder_multimodal llm_engine_example load_sharded_state lora_with_quantization_inference mistral-small mlpspeculator multilora_inference neuron neuron_int8_quantization openai prefix_caching prithvi_geospatial_mae profiling profiling_tpu reproduciblity rlhf rlhf_colocate rlhf_utils save_sharded_state simple_profiling structured_outputs torchrun_example tpu vision_language vision_language_embedding vision_language_multi_image :::