Available Pipelines

All pipelines can be run with:

uv run parse-bench run <pipeline_name>

To see the full list:

uv run parse-bench pipelines

Setup

Copy .env.example to .env and fill in the API keys / endpoints for the providers you want to use:

cp .env.example .env

Cloud API Pipelines

These pipelines use hosted APIs. You only need an API key in your .env file.

Bold pipelines are baselines evaluated in the ParseBench paper. The name used in the paper is shown in parentheses.

Upstage

Pipeline	Description	Env Var
`upstage_dpe_v2`	Document Parse 2 (Enhanced), snapshot 260930	`UPSTAGE_API_KEY`

Install with uv sync --extra upstage. PDFs are uploaded unchanged. Native images smaller than a 2000px short edge are proportionally upscaled; larger images retain their dimensions. The pipeline requests HTML, Markdown, text and coordinates. PDF rendering uses API defaults. The provider uses API Markdown for body text, retaining HTML for tables, charts, figures and lists. It preserves OCR text and authored table captions, removes generated image narration and transport attributes from evaluable markup, and converts closed HTML code blocks to fences. Literal code, table spans and layout coordinates are preserved.

uv run parse-bench run upstage_dpe_v2 --test --max_concurrent 2

LlamaParse

Pipeline	Description	Env Var
`llamaparse_agentic`	Agentic tier (In paper: LlamaParse Agentic)	`LLAMA_CLOUD_API_KEY`
`llamaparse_cost_effective`	Cost-effective tier (In paper: LlamaParse Cost Effective)	`LLAMA_CLOUD_API_KEY`
`llamaparse_agentic_plus`	Agentic plus tier	`LLAMA_CLOUD_API_KEY`

OpenAI

Pipeline	Description	Env Var
`openai_gpt_6_1_sol_reasoning_low_parse_with_layout_file`	GPT-6.1 Sol, low reasoning + layout, PDF file mode	`OPENAI_API_KEY`
`openai_gpt_6_1_terra_reasoning_low_parse_with_layout_file`	GPT-6.1 Terra, low reasoning + layout, PDF file mode	`OPENAI_API_KEY`
`openai_gpt_6_1_terra_reasoning_med_parse_with_layout_file`	GPT-6.1 Terra, medium reasoning + layout, PDF file mode	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_medium_parse`	GPT-5 Mini, medium reasoning, image mode	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_medium_parse_file`	GPT-5 Mini, medium reasoning, PDF file mode	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_minimal_parse`	GPT-5 Mini, minimal reasoning	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_minimal_parse_file`	GPT-5 Mini, minimal reasoning, file mode	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_medium_parse_with_layout`	GPT-5 Mini, medium reasoning + layout	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_medium_parse_with_layout_file`	GPT-5 Mini, medium reasoning + layout, file (In paper: OpenAI GPT-5 Mini (Reasoning Medium))	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_minimal_parse_with_layout`	GPT-5 Mini, minimal reasoning + layout	`OPENAI_API_KEY`
`openai_gpt5_mini_reasoning_minimal_parse_with_layout_file`	GPT-5 Mini, minimal reasoning + layout, file (In paper: OpenAI GPT-5 Mini (Reasoning Minimal))	`OPENAI_API_KEY`
`openai_gpt_5_4_parse`	GPT-5.4, image mode	`OPENAI_API_KEY`
`openai_gpt_5_4_parse_file`	GPT-5.4, PDF file mode	`OPENAI_API_KEY`
`openai_gpt_5_4_reasoning_none_parse`	GPT-5.4, reasoning disabled, image mode	`OPENAI_API_KEY`
`openai_gpt_5_4_reasoning_none_parse_file`	GPT-5.4, reasoning disabled, PDF file mode	`OPENAI_API_KEY`
`openai_gpt_5_4_parse_with_layout_file`	GPT-5.4, parse + layout, file mode (In paper: OpenAI GPT-5.4)	`OPENAI_API_KEY`

Anthropic Claude

Pipeline	Description	Env Var
`anthropic_haiku_5_5_parse_with_layout_file`	Claude Haiku 5.5, adaptive thinking + layout, PDF file mode	`ANTHROPIC_API_KEY`
`anthropic_haiku_parse`	Claude Haiku 4.5, image mode	`ANTHROPIC_API_KEY`
`anthropic_haiku_parse_file`	Claude Haiku 4.5, PDF file mode	`ANTHROPIC_API_KEY`
`anthropic_haiku_parse_with_layout`	Claude Haiku 4.5, parse + layout	`ANTHROPIC_API_KEY`
`anthropic_haiku_parse_with_layout_file`	Claude Haiku 4.5, parse + layout, file mode (In paper: Anthropic Haiku 4.5 (Disable Thinking))	`ANTHROPIC_API_KEY`
`anthropic_haiku_thinking_parse_with_layout_file`	Claude Haiku 4.5, extended thinking + layout (In paper: Anthropic Haiku 4.5 (Thinking))	`ANTHROPIC_API_KEY`
`anthropic_opus_4_6_parse`	Claude Opus 4.6, image mode	`ANTHROPIC_API_KEY`
`anthropic_opus_4_6_parse_file`	Claude Opus 4.6, PDF file mode	`ANTHROPIC_API_KEY`
`anthropic_opus_4_6_parse_with_layout_file`	Claude Opus 4.6, parse + layout, file mode (In paper: Anthropic Opus 4.6)	`ANTHROPIC_API_KEY`
`anthropic_opus_4_8_parse_with_layout_file`	Claude Opus 4.8, parse + layout, file mode (In paper: Anthropic Opus 4.8)	`ANTHROPIC_API_KEY`
`anthropic_sonnet_5_parse_with_layout`	Claude Sonnet 5, adaptive thinking + layout, image mode	`ANTHROPIC_API_KEY`
`anthropic_sonnet_5_parse_with_layout_file`	Claude Sonnet 5, adaptive thinking + layout, file mode	`ANTHROPIC_API_KEY`
`anthropic_fable_5_parse_with_layout_file`	Claude Fable 5, parse + layout, file mode	`ANTHROPIC_API_KEY`
`anthropic_fable_5_1_parse_with_layout_file`	Claude Fable 5.1, parse + layout, file mode	`ANTHROPIC_API_KEY`

Google Gemini

Pipeline	Description	Env Var
`google_gemini_3_flash_lite_parse`	Gemini 3 Flash Lite, image mode	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_lite_parse_file`	Gemini 3 Flash Lite, file mode	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_minimal_parse`	Gemini 3 Flash, minimal thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_minimal_parse_file`	Gemini 3 Flash, minimal thinking, file	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_high_parse`	Gemini 3 Flash, high thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_high_parse_file`	Gemini 3 Flash, high thinking, file	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_minimal_parse_with_layout`	Gemini 3 Flash, minimal thinking + layout	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_high_parse_with_layout`	Gemini 3 Flash, high thinking + layout	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_minimal_parse_with_layout_file`	Gemini 3 Flash, minimal thinking + layout file (In paper: Google Gemini 3 Flash (Thinking Minimal))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_high_parse_with_layout_file`	Gemini 3 Flash, high thinking + layout file (In paper: Google Gemini 3 Flash (Thinking High))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_minimal_parse_with_layout_agentic_vision`	Agentic vision, minimal thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_medium_parse_with_layout_agentic_vision`	Agentic vision, medium thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_flash_thinking_high_parse_with_layout_agentic_vision`	Agentic vision, high thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_1_flash_lite_parse`	Gemini 3.1 Flash Lite	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_1_flash_lite_thinking_high_parse`	Gemini 3.1 Flash Lite, high thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_1_flash_lite_thinking_high_parse_with_layout_file`	Gemini 3.1 Flash Lite, high thinking + layout, file mode	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_1_pro_parse`	Gemini 3.1 Pro, default thinking	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_1_pro_parse_with_layout_file`	Gemini 3.1 Pro, parse + layout, file mode (In paper: Google Gemini 3.1 Pro)	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_5_flash_parse_with_layout`	Gemini 3.5 Flash, default thinking + layout	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_5_flash_no_thinking_parse_with_layout`	Gemini 3.5 Flash, minimal thinking + layout	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_5_flash_parse_with_layout_file`	Gemini 3.5 Flash, default thinking + layout, file mode (In paper: Google Gemini 3.5 Flash (Thinking Medium))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_5_flash_no_thinking_parse_with_layout_file`	Gemini 3.5 Flash, minimal thinking + layout, file mode (In paper: Google Gemini 3.5 Flash (Thinking Minimal))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_5_flash_lite_parse_with_layout_file`	Gemini 3.5 Flash Lite, layout + file mode (In paper: Google Gemini 3.5 Flash Lite)	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_6_flash_parse_with_layout_file`	Gemini 3.6 Flash, default thinking + layout, file mode (In paper: Google Gemini 3.6 Flash (Thinking Medium))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_6_flash_no_thinking_parse_with_layout_file`	Gemini 3.6 Flash, minimal thinking + layout, file mode (In paper: Google Gemini 3.6 Flash (Thinking Minimal))	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_6_flash_parse_with_layout`	Gemini 3.6 Flash, default thinking + layout, image mode	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_6_flash_no_thinking_parse_with_layout`	Gemini 3.6 Flash, minimal thinking + layout, image mode	`GOOGLE_GEMINI_API_KEY`
`google_gemini_3_7_flash_thinking_high_parse_with_layout_file`	Gemini 3.7 Flash, high thinking + layout, file mode	`GOOGLE_GEMINI_API_KEY`

GLM (z.ai)

Pipeline	Description	Env Var
`glm_5_3_flash_parse_with_layout_file`	GLM-5.3 Flash, parse + layout, file mode	`GLM_ZAI_API_KEY`
`deepseek_v4_1_flash_parse_with_layout`	DeepSeek-V4.1-Flash, parse + layout, thinking high	`DEEPSEEK_API_KEY`
`deepseek_v4_1_flash_no_thinking_parse_with_layout`	DeepSeek-V4.1-Flash, parse + layout, thinking disabled	`DEEPSEEK_API_KEY`

Azure Document Intelligence

Pipeline	Description	Env Vars
`azure_di_layout`	Layout model (In paper: Azure Document Intelligence)	`AZURE_DOCUMENT_INTELLIGENCE_KEY`, `AZURE_DOCUMENT_INTELLIGENCE_ENDPOINT`
`azure_di_read`	Read model	`AZURE_DOCUMENT_INTELLIGENCE_KEY`, `AZURE_DOCUMENT_INTELLIGENCE_ENDPOINT`

AWS Textract

Pipeline	Description	Env Vars
`aws_textract`	Standard Textract (In paper: AWS Textract)	`AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`
`aws_textract_with_forms`	Textract with forms	`AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`
`aws_textract_text_only`	Textract text only	`AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`

Google Document AI

Pipeline	Description	Env Vars
`google_docai`	Document AI OCR	`GOOGLE_DOCAI_PROJECT_ID`, `GOOGLE_DOCAI_PROCESSOR_ID`
`google_docai_layout`	Document AI Layout (In paper: Google Cloud Document AI)	`GOOGLE_DOCAI_PROJECT_ID`, `GOOGLE_DOCAI_LAYOUT_PROCESSOR_ID`

Reducto

Pipeline	Description	Env Var
`reducto`	Default Reducto (In paper: Reducto)	`REDUCTO_API_KEY`
`reducto_agentic`	Agentic mode (In paper: Reducto (Agentic))	`REDUCTO_API_KEY`
`reducto_r1`	r-1 model, `settings.model="r-1"` (In paper: Reducto (r-1))	`REDUCTO_API_KEY`
`reducto_nonagentic_change_tracking`	Non-agentic with `formatting_include=["change_tracking"]`	`REDUCTO_API_KEY`
`reducto_agentic_table`	Agentic on text + table scopes	`REDUCTO_API_KEY`
`reducto_agentic_chart`	Agentic on text + table + figure scopes with advanced chart agent	`REDUCTO_API_KEY`

Mistral OCR

Pipeline	Description	Env Var
`mistral_ocr_4`	Mistral OCR 4.0	`MISTRAL_API_KEY`
`mistral_ocr_4_annotation`	Mistral OCR 4.0 with bbox annotation (Document AI)	`MISTRAL_API_KEY`
`mistral_ocr_4_1`	Mistral OCR 4.1	`MISTRAL_API_KEY`
`mistral_ocr_4_1_annotation`	Mistral OCR 4.1 with bbox annotation (Document AI)	`MISTRAL_API_KEY`

Pulse

Pipeline	Description	Env Var
`pulse_ultra_2`	`pulse-ultra-2` hosted tier with native markdown output, refinement enabled, and inline formatting preserved	`PULSE_API_KEY`

Chunkr

Pipeline	Description	Env Var
`chunkr`	Default quality	`CHUNKR_API_KEY`
`chunkr_high_res`	High resolution	`CHUNKR_API_KEY`

Datalab (Marker)

Pipeline	Description	Env Var
`datalab_fast`	Fast mode	`DATALAB_API_KEY`
`datalab_balanced`	Balanced mode	`DATALAB_API_KEY`
`datalab_accurate`	Accurate mode	`DATALAB_API_KEY`

Extend AI

Pipeline	Description	Env Var
`extend_parse`	Default (In paper: Extend)	`EXTEND_API_KEY`
`extend_parse_2`	2.0 engine (v2.0.0, GA)	`EXTEND_API_KEY`
`extend_parse_light`	Light engine (v1.0.0)	`EXTEND_API_KEY`
`extend_parse_document`	Document scope	`EXTEND_API_KEY`
`extend_parse_section`	Section scope	`EXTEND_API_KEY`

Landing AI

Pipeline	Description	Env Var
`landingai_parse`	Default (In paper: LandingAI)	`LANDING_AI_API_KEY`

Unstructured

Pipeline	Description	Env Var
`unstructured_auto`	Auto strategy	`UNSTRUCTURED_API_KEY`
`unstructured_fast`	Fast strategy	`UNSTRUCTURED_API_KEY`
`unstructured_hi_res`	Hi-res strategy	`UNSTRUCTURED_API_KEY`

OpenInnovation Parser (oi-parser)

Hosted document-parsing API. Sign up at oi-parser.ai to get an API key.

Pipeline	Description	Env Vars
`oi_parser`	oi-parser hosted `/v1/extract` API	`OI_PARSER_API_KEY`, `OI_PARSER_BASE_URL` (optional)

DocAI (ProvidusAI)

Hosted parse API. Documents are uploaded into a knowledge base named ParseBench (created on first use) with auto-parse on; the job is polled and result.md plus grounding.json are fetched. Tables come back as HTML with row and column spans; grounding boxes feed the layout vertical through the docai layout adapter. Cost is one credit per parsed page at the pay-as-you-go rate of USD 0.01 (1¢ per page); override with credit_rate_usd or DOCAI_CREDIT_RATE_USD. Sign up at https://platform.providus.ai. Recommended --max_concurrent 30.

Pipeline	Description	Env Vars
`docai_default`	DocAI parsing pipeline by ProvidusAI (In paper: DocAI)	`DOCAI_API_KEY`, `DOCAI_BASE_URL` (optional)

anyformat

Hosted parse API. Each pipeline creates a one-node parse workflow for its tier (cache off) on first use, runs every document through POST /v3/workflows/{id}/upload/run/ and polls GET /v3/runs/{id}/. Cost is the public list price of the tier (credits per page × credit rate); the API reports no usage. Layout bboxes come from the response's blocks[] and feed the layout vertical through the anyformat layout adapter.

Pipeline	Description	Env Vars
`anyformat_standard`	Standard tier (25 credits/page)	`ANYFORMAT_API_KEY`, `ANYFORMAT_BASE_URL` (optional)
`anyformat_agentic`	Agentic tier (100 credits/page)	same
`anyformat_lite`	Fast tier (12 credits/page)	same
`anyformat_flash`	Flash tier (7 credits/page)	same

Config keys accepted through PipelineSpec.config: mode, workflow_id (reuse an existing one-node parse workflow), effort, prompt_hint, figure_enhancement, credit_rate_usd, poll_interval, request_timeout, job_timeout. Recommended --max_concurrent 10 (submission tier: 60 requests/min).

Nutrient DWS (Data Extraction API)

Pipeline	Description	Env Var
`nutrient_dws_text`	`text` mode, markdown only (no spatial output)	`NUTRIENT_DWS_API_KEY`
`nutrient_dws_structure`	`structure` mode, OCR-backed with spatial elements	`NUTRIENT_DWS_API_KEY`
`nutrient_dws_understand`	`understand` mode, adds semantic enrichment	`NUTRIENT_DWS_API_KEY`
`nutrient_dws_agentic`	`agentic` mode, visual reasoning (In paper: Nutrient DWS (Agentic))	`NUTRIENT_DWS_API_KEY`

Self-hosted Model Pipelines

These pipelines require you to deploy the model on your own infrastructure (e.g., via vLLM, Modal, etc.) and set the endpoint URL in .env.

Gemma 4

Pipeline	Description	Env Var
`gemma4_26b_vllm`	Gemma 4 26B-A4B, parse mode	`GEMMA4_SERVER_URL`
`gemma4_26b_vllm_with_layout`	Gemma 4 26B-A4B, layout mode	`GEMMA4_SERVER_URL`
`gemma4_e4b_vllm`	Gemma 4 E4B (dense 8B), parse mode	`GEMMA4_SERVER_URL`
`gemma4_e4b_vllm_with_layout`	Gemma 4 E4B, layout mode	`GEMMA4_SERVER_URL`

Qwen

Pipeline	Description	Env Var
`qwen3_5_4b_vllm_parse`	Parse mode, markdown (In paper: Qwen 3 VL)	`QWEN35_SERVER_URL`
`qwen3_5_4b_vllm_layout`	Layout mode, JSON with bboxes (In paper: Qwen 3 VL)	`QWEN35_SERVER_URL`
`qwen3_8_27b_parse_with_layout`	Qwen3.8-27B-FP8, parse + layout	`QWEN3_8_27B_SERVER_URL`
`qwen3_8_27b_thinking_parse_with_layout`	Qwen3.8-27B-FP8, parse + layout with thinking	`QWEN3_8_27B_SERVER_URL`
`qwen3_8_flash_next_parse_with_layout`	Qwen3.8-Flash-Next-FP8, parse + layout	`QWEN3_8_FLASH_NEXT_SERVER_URL`
`qwen3_8_flash_next_thinking_parse_with_layout`	Qwen3.8-Flash-Next-FP8, parse + layout with thinking	`QWEN3_8_FLASH_NEXT_SERVER_URL`

LightOnOCR-3

Pipeline	Description	Env Var
`lightonocr_3_vllm_parse`	LightOnOCR-3 grounding prompt, parse + layout from one pass (400 DPI, 5M pixel cap)	`LIGHTONOCR_3_SERVER_URL`

Chandra OCR 2

Pipeline	Description	Env Var
`chandra2_vllm`	OpenAI-compatible vLLM API	`CHANDRA2_SERVER_URL`
`chandra2_sdk`	Official SDK endpoint	`CHANDRA2_SERVER_URL`

DeepSeek-OCR-2

Pipeline	Description	Env Var
`deepseekocr2_vllm`	With grounding layout detection	`DEEPSEEKOCR2_SERVER_URL`
`deepseekocr2_freeocr`	Free OCR, no grounding	`DEEPSEEKOCR2_SERVER_URL`

Granite Vision

Pipeline	Description	Env Var
`granite_vision_pipeline`	PP-DocLayout + per-region Granite Vision	`GRANITE_VISION_SERVER_URL`
`granite_vision_4_1_4b`	Granite Vision 4.1 4B (vLLM, multi-task)	`VLLM_API_KEY`

PaddleOCR-VL

Pipeline	Description	Env Var
`paddleocr_vl_vllm`	OpenAI-compatible vLLM API	`PADDLEOCR_SERVER_URL`
`paddleocr_vl_pipeline`	Full pipeline (layout + chart routing)	`PADDLEOCR_SERVER_URL`
`paddleocr_vl_1_6_vllm`	PaddleOCR-VL-1.6, OCR prompt	`PADDLEOCR_SERVER_URL`
`paddleocr_vl_1_6_vllm_table`	PaddleOCR-VL-1.6, table recognition prompt	`PADDLEOCR_SERVER_URL`
`paddleocr_vl_1_6_pipeline`	PaddleOCR-VL-1.6, full pipeline (layout + routing)	`PADDLEOCR_SERVER_URL`

dots.ocr

Pipeline	Description	Env Var
`dots_ocr_1_0_parse`	dots.ocr 1.0	`DOTS_OCR_ENDPOINT_URL`
`dots_ocr_1_5_parse`	dots.ocr 1.5, layout+text prompt (In paper: Dots OCR 1.5)	`DOTS_OCR_ENDPOINT_URL`

Docling

Pipeline	Description	Env Vars
`docling_parse`	Docling HTTP endpoint (In paper: Docling)	`DOCLING_PARSE_ENDPOINT_URL`, `DOCLING_PARSE_API_KEY` (optional)
`docling_serve`	Docling Serve HTTP endpoint	`DOCLING_SERVE_ENDPOINT_URL`, `DOCLING_SERVE_API_KEY` (optional)

MinerU 2.5

Pipeline	Description	Env Var
`mineru25_vllm`	MinerU2.5-2509-1.2B vLLM server (two-step layout + recognition)	`MINERU25_SERVER_URL`
`mineru2605pro_vllm`	MinerU2.5-Pro-2605-1.2B vLLM server (adds chart/image analysis)	`MINERU2605PRO_SERVER_URL`

MinerU-Diffusion

Pipeline	Description	Env Var
`mineru_diffusion`	MinerU-Diffusion-V1-0320-2.5B server (diffusion-decoding OCR, two-stage layout + recognition)	`MINERU_DIFFUSION_SERVER_URL`

Nemotron-Omni

Pipeline	Description	Env Var
`nemotron_omni_30b_vllm`	Nemotron-3-Nano-Omni 30B-A3B Reasoning, thinking disabled	`NEMOTRON_OMNI_SERVER_URL`
`nemotron_omni_30b_vllm_thinking`	Nemotron-3-Nano-Omni 30B-A3B Reasoning, thinking enabled	`NEMOTRON_OMNI_SERVER_URL`

Nemotron 3.5 Parse

Pipeline	Description	Env Var
`nemotron_parse_3_5_sdk`	Nemotron 3.5 Parse SDK server (triple pass: full page + detection + Table/Chart crops; layout)	`NEMOTRON_PARSE_3_5_SERVER_URL`

MiniCPM-OCR 1.0

Pipeline	Description	Env Var
`minicpm_ocr_sdk`	MiniCPM-OCR 1.0 `minicpmvocr` SDK server (PP-DocLayoutV3 regions recognized per label; layout)	`MINICPM_OCR_SERVER_URL`

Unlimited-OCR 2

Pipeline	Description	Env Var
`unlimitedocr2`	Unlimited-OCR 2 on its official SGLang server (whole document per request; layout)	`UNLIMITEDOCR2_SERVER_URL`

OvisOCR3

Pipeline	Description	Env Var
`ovisocr3_vllm`	OvisOCR3 on vLLM with the model card settings (official structured parse plus a chart-parsing pass per chart region; layout)	`OVISOCR3_SERVER_URL`

Surya OCR 2

Pipeline	Description	Env Var
`surya2_sdk`	Surya OCR 2 SDK server (full-page OCR + layout)	`SURYA2_SERVER_URL`

florin-parser-nano

Fine-tune of KDL-Frontier-Parser-nano (florin-inc/florin-parser-nano); served identically to the base model and driven by the same two-stage pipeline, with markdown-emission fixes (see the provider module docstring for the exact vllm serve command).

Pipeline	Description	Env Var
`florin_parser_nano`	vLLM OpenAI-compatible endpoint (layout + per-region recognition)	`FLORIN_NANO_ENDPOINT_URL`

rakedoc-nano

Fine-tune of florin-parser-nano (cloudraker/rakedoc-nano); served identically to the parent models and driven by the same two-stage pipeline and markdown emission (see the provider module docstring for the exact vllm serve command).

Pipeline	Description	Env Var
`rakedoc_nano`	vLLM OpenAI-compatible endpoint (layout + per-region recognition)	`RAKEDOC_NANO_ENDPOINT_URL`

aeddix-alpine-ocr (KDL pipeline)

Fine-tune of MinerU2.5-Pro-2605-1.2B (aeddix-labs/aeddix-alpine-ocr, tag v0.1); served identically to KDL-Frontier-Parser-nano and driven by the unchanged kdl_frontier_nano two-stage pipeline and markdown emission (see the provider module docstring for the exact vllm serve command).

Pipeline	Description	Env Var
`aeddix_alpine_ocr_kdl`	vLLM OpenAI-compatible endpoint (layout + per-region recognition)	`AEDDIX_ALPINE_OCR_KDL_ENDPOINT_URL`

Local Pipelines (No API key needed)

These run entirely locally and do not require API keys.

Pipeline	Description	Requirements
`pypdf_baseline`	PyPDF text extraction	None
`pymupdf_text`	PyMuPDF text extraction	None
`pymupdf_html`	PyMuPDF HTML extraction	None
`liteparse_markdown`	LiteParse Markdown (no OCR; layout blocks via `--extract-blocks`; the leaderboard's LiteParse (no OCR) row)	`pip install "parse-bench[liteparse]"` (or `LITEPARSE_BIN=/path/to/lit`)
`liteparse_text`	LiteParse plain text (no OCR)	same as above
`pymupdf4llm_markdown`	PyMuPDF4LLM Markdown with native HTML tables and RapidOCR at 150 DPI	Python 3.12, `pymupdf==2.1`, and `rapidocr==3.10.0`
`warp_ingest`	Warp-Ingest local parser	`warp-ingest[ocr]>=2.0.1` installed
`tesseract_eng`	Tesseract OCR (English)	`tesseract` installed
`tesseract_fast`	Tesseract OCR (fast)	`tesseract` installed
`tesseract_high_quality`	Tesseract OCR (high quality)	`tesseract` installed
`infinity_parser2_flash`	Infinity-Parser2-Flash (vLLM server, JSON layout)	`infinity_parser2`, vLLM server at `INFINITY_PARSER2_API_URL`
`infinity_parser2_pro`	Infinity-Parser2-Pro (vLLM server, JSON layout)	`infinity_parser2`, vLLM server at `INFINITY_PARSER2_API_URL`

Create the isolated Python environment for pymupdf4llm_markdown with uv sync --python 3.12 --extra pymupdf4llm, then run it with uv run --python 3.12 --extra pymupdf4llm parse-bench run pymupdf4llm_markdown --max_concurrent 1.

Layout Detection Pipelines

Pipeline	Description	Requirements
`docling_layout_heron`	Docling Heron layout	Self-hosted endpoint
`docling_layout_heron_101`	Docling Heron 1.0.1	Self-hosted endpoint
`docling_layout_old`	Docling legacy layout	Self-hosted endpoint
`ppdoclayout_plus_l`	PaddleDetection layout	Self-hosted endpoint
`qwen3vl_layout`	Qwen3-VL layout	Self-hosted endpoint
`surya_layout`	Surya layout detection	`surya` installed
`yolo_doclaynet`	YOLO DocLayNet	Self-hosted endpoint

FilesExpand file tree

pipelines.md

Latest commit

History