News

Benchmark

Installation

sudo apt-get updatesudo apt-get install poppler-utils ttf-mscorefonts-installer msttcorefonts fonts-crosextra-caladea fonts-crosextra-carlito gsfonts lcdf-typetools

conda create -n olmocr python=3.11conda activate olmocr# For CPU-only operations, ex running the benchmarkpip install olmocr[bench]# For actually converting the files with your own GPUpip install olmocr[gpu]  --extra-index-url https://download.pytorch.org/whl/cu128# Recommended: Install flash infer for faster inference on GPUpip install https://download.pytorch.org/whl/cu128/flashinfer/flashinfer_python-0.2.5%2Bcu128torch2.7-cp38-abi3-linux_x86_64.whl

Local Usage Example

# Download a sample PDFcurl -o olmocr-sample.pdf https://olmocr.allenai.org/papers/olmocr_3pg_sample.pdf# Convert it to markdownpython -m olmocr.pipeline ./localworkspace --markdown --pdfs olmocr-sample.pdf

python -m olmocr.pipeline ./localworkspace --markdown --pdfs random_page.png

python -m olmocr.pipeline ./localworkspace --markdown --pdfs tests/gnarly_pdfs/*.pdf

Viewing Results

cat localworkspace/markdown/olmocr-sample.md

olmOCR: Unlocking Trillions of Tokens in PDFs with Vision Language Models...

Using an Inference Provider or External Server

# Use external vLLM server instead of local onepython -m olmocr.pipeline ./localworkspace --server http://remote-server:8000/v1 --markdown --pdfs tests/gnarly_pdfs/*.pdf

vllm serve allenai/olmOCR-2-7B-1025-FP8 --served-model-name olmocr --max-model-len 16384

Verified External Providers

Multi-node / Cluster Usage

python -m olmocr.pipeline s3://my_s3_bucket/pdfworkspaces/exampleworkspace --pdfs s3://my_s3_bucket/jakep/gnarly_pdfs/*.pdf

python -m olmocr.pipeline s3://my_s3_bucket/pdfworkspaces/exampleworkspace

python -m olmocr.pipeline s3://my_s3_bucket/pdfworkspaces/exampleworkspace --pdfs s3://my_s3_bucket/jakep/gnarly_pdfs/*.pdf --beaker --beaker_gpus 4

Using Docker

docker pull alleninstituteforai/olmocr:latest

docker run -it --gpus all --name olmocr_container alleninstituteforai/olmocr:latest /bin/bash

docker run -it --gpus all \  -v /path/to/your/local/files:/local_files \  --name olmocr_container \  alleninstituteforai/olmocr:latest /bin/bash

curl -o olmocr-sample.pdf https://olmocr.allenai.org/papers/olmocr_3pg_sample.pdfpython -m olmocr.pipeline ./localworkspace --markdown --pdfs olmocr-sample.pdf

Full documentation for the pipeline

python -m olmocr.pipeline --helpusage: pipeline.py [-h] [--pdfs [PDFS ...]] [--model MODEL] [--workspace_profile WORKSPACE_PROFILE] [--pdf_profile PDF_PROFILE] [--pages_per_group PAGES_PER_GROUP] [--max_page_retries MAX_PAGE_RETRIES] [--max_page_error_rate MAX_PAGE_ERROR_RATE] [--workers WORKERS]                   [--apply_filter] [--stats] [--markdown] [--target_longest_image_dim TARGET_LONGEST_IMAGE_DIM] [--target_anchor_text_len TARGET_ANCHOR_TEXT_LEN] [--guided_decoding] [--gpu-memory-utilization GPU_MEMORY_UTILIZATION] [--max_model_len MAX_MODEL_LEN]                   [--tensor-parallel-size TENSOR_PARALLEL_SIZE] [--data-parallel-size DATA_PARALLEL_SIZE] [--port PORT] [--server SERVER] [--beaker] [--beaker_workspace BEAKER_WORKSPACE] [--beaker_cluster BEAKER_CLUSTER] [--beaker_gpus BEAKER_GPUS] [--beaker_priority BEAKER_PRIORITY]                   workspaceManagerfor running millions of PDFs through a batch inference pipelinepositional arguments:  workspace             The filesystem path where work will be stored, can be alocal folder, or an s3 pathif coordinating work with many workers, s3://bucket/prefix/options:  -h, --help            show thishelp message andexit  --pdfs [PDFS ...]     Path to add pdfs storedin s3 to the workspace, can be a glob path s3://bucket/prefix/*.pdf or path to file containing list of pdf paths  --model MODEL         Path where the model is located, allenai/olmOCR-7B-0725-FP8 is the default, can be local, s3, or hugging face.  --workspace_profile WORKSPACE_PROFILE                        S3 configuration profilefor accessing the workspace  --pdf_profile PDF_PROFILE                        S3 configuration profilefor accessing the raw pdf documents  --pages_per_group PAGES_PER_GROUP                        Aimingfor this many pdf pages per work item group  --max_page_retries MAX_PAGE_RETRIES                        Max number oftimes we will retry rendering a page  --max_page_error_rate MAX_PAGE_ERROR_RATE                        Rate of allowable failed pagesin a document, 1/250 by default  --workers WORKERS     Number of workers to run at atime  --apply_filter        Apply basic filtering to English pdfs which are not forms, and not likely seo spam  --stats               Instead of running any job, reports some statistics about the current workspace  --markdown            Also write natural text to markdown files preserving the folder structure of the input pdfs  --target_longest_image_dim TARGET_LONGEST_IMAGE_DIM                        Dimension on longest side to usefor rendering the pdf pages  --target_anchor_text_len TARGET_ANCHOR_TEXT_LEN                        Maximum amount of anchor text to use (characters), not usedfor new models  --guided_decoding     Enable guided decodingfor model YAMLtype outputsVLLM arguments:  --gpu-memory-utilization GPU_MEMORY_UTILIZATION                        Fraction of VRAM vLLM may pre-allocatefor KV-cache (passed through to vllm serve).  --max_model_len MAX_MODEL_LEN                        Upper bound (tokens) vLLM will allocate KV-cache for, lowerif VLLM won't start  --tensor-parallel-size TENSOR_PARALLEL_SIZE, -tp TENSOR_PARALLEL_SIZE                        Tensor parallel size for vLLM  --data-parallel-size DATA_PARALLEL_SIZE, -dp DATA_PARALLEL_SIZE                        Data parallel size for vLLM  --port PORT           Port to use for the VLLM server  --server SERVER       URL of external vLLM (or other compatible provider)                        server (e.g., http://hostname:port). If provided,                        skips spawning local vLLM instancebeaker/cluster execution:  --beaker              Submit this job to beaker instead of running locally  --beaker_workspace BEAKER_WORKSPACE                        Beaker workspace to submit to  --beaker_cluster BEAKER_CLUSTER                        Beaker clusters you want to run on  --beaker_gpus BEAKER_GPUS                        Number of gpu replicas to run  --beaker_priority BEAKER_PRIORITY                        Beaker priority level for the job

Code overview

Team

License

Citing

@misc{olmocrbench,title={{olmOCR: Unlocking Trillions of Tokens in PDFs with Vision Language Models}},author={Jake Poznanski and Jon Borchardt and Jason Dunkelberger and Regan Huff and Daniel Lin and Aman Rangapur and Christopher Wilhelm and Kyle Lo and Luca Soldaini},year={2025},eprint={2502.18443},archivePrefix={arXiv},primaryClass={cs.CL},url={https://arxiv.org/abs/2502.18443},}

@misc{olmocr2,title={olmOCR 2: Unit Test Rewards for Document OCR},author={Jake Poznanski and Luca Soldaini and Kyle Lo},year={2025},eprint={2510.19817},archivePrefix={arXiv},primaryClass={cs.CV},url={https://arxiv.org/abs/2510.19817}, }