{ "hf_id": "PaddlePaddle/PaddleOCR-VL-1.6", "meta": { "title": "PaddleOCR-VL-1.6", "slug": "paddleocr-vl-1.6", "provider": "PaddlePaddle", "description": "PaddleOCR-VL-1.6 (0.9B) — region-aware data optimization + progressive post-training; new SOTA 96.33% on OmniDocBench v1.6, drop-in replacement for 1.5", "date_added": "2026-07-11", "date_updated": "2026-07-08", "tasks": [ "multimodal" ], "related_recipes": [ "PaddlePaddle/PaddleOCR-VL-1.5", "PaddlePaddle/PaddleOCR-VL" ] }, "recommended_command": { "hardware": "h200", "strategy": "single_node_tp", "variant": "default", "node_count": 1, "deploy_type": "single_node", "env": {}, "docker_image": "vllm/vllm-openai:latest", "command": "vllm serve PaddlePaddle/PaddleOCR-VL-1.6 \\\n --trust-remote-code \\\n --max-num-batched-tokens 16384 \\\n --no-enable-prefix-caching \\\n --mm-processor-cache-gb 0 \\\n --tensor-parallel-size 1", "argv": [ "vllm", "serve", "PaddlePaddle/PaddleOCR-VL-1.6", "--trust-remote-code", "--max-num-batched-tokens", "16384", "--no-enable-prefix-caching", "--mm-processor-cache-gb", "0", "--tensor-parallel-size", "1" ], "docker_command": "docker run --gpus all \\\n --privileged --ipc=host -p 8000:8000 \\\n -v ~/.cache/huggingface:/root/.cache/huggingface \\\n vllm/vllm-openai:latest PaddlePaddle/PaddleOCR-VL-1.6 \\\n --trust-remote-code \\\n --max-num-batched-tokens 16384 \\\n --no-enable-prefix-caching \\\n --mm-processor-cache-gb 0 \\\n --tensor-parallel-size 1", "docker_argv": [ "docker", "run", "--gpus", "all", "--privileged", "--ipc=host", "-p", "8000:8000", "-v", "~/.cache/huggingface:/root/.cache/huggingface", "vllm/vllm-openai:latest", "PaddlePaddle/PaddleOCR-VL-1.6", "--trust-remote-code", "--max-num-batched-tokens", "16384", "--no-enable-prefix-caching", "--mm-processor-cache-gb", "0", "--tensor-parallel-size", "1" ], "strategy_spec": { "name": "single_node_tp", "deploy_type": "single_node", "display_name": "Tensor Parallel", "short_label": "TP", "orientation": "latency", "description": "Single-node tensor parallel. Splits the model across all local GPUs. TP size is set to the GPU count at deploy time. The simplest multi-GPU strategy — works for all model architectures.\n", "hardware_match": { "min_gpus": 1, "max_gpus": 8, "multi_node": false }, "vllm_args": [], "parallel_flag": "--tensor-parallel-size" }, "hardware_profile": { "brand": "NVIDIA", "generation": "hopper", "compute_arch": "sm90", "display_name": "H200", "description": "NVIDIA H200 SXM 141 GB HBM3e · 8-GPU HGX node", "gpu_count": 8, "vram_gb": 1128, "multi_node": false }, "alternatives": { "multi_node_tp": "/PaddlePaddle/PaddleOCR-VL-1.6/strategies/multi_node_tp.json", "kv_store_centralized_mooncake": "/PaddlePaddle/PaddleOCR-VL-1.6/strategies/kv_store_centralized_mooncake.json", "kv_store_distributed_mooncake": "/PaddlePaddle/PaddleOCR-VL-1.6/strategies/kv_store_distributed_mooncake.json" }, "by_hardware": { "h200": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/h200.json", "h100": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/h100.json", "b200": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/b200.json", "gb200": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/gb200.json", "b300": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/b300.json", "gb300": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/gb300.json", "mi300x": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/mi300x.json", "mi325x": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/mi325x.json", "mi350x": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/mi350x.json", "mi355x": "/PaddlePaddle/PaddleOCR-VL-1.6/hw/mi355x.json" } }, "model": { "model_id": "PaddlePaddle/PaddleOCR-VL-1.6", "min_vllm_version": "0.11.1", "architecture": "dense", "parameter_count": "0.9B", "active_parameters": "0.9B", "context_length": 131072, "base_args": [ "--trust-remote-code", "--max-num-batched-tokens", "16384", "--no-enable-prefix-caching", "--mm-processor-cache-gb", "0" ], "base_env": {}, "install": { "pip": { "command": "uv venv\nsource .venv/bin/activate\nuv pip install -U vllm --torch-backend auto" }, "docker": { "command": "docker pull vllm/vllm-openai:latest" }, "extras": [ { "note": "PaddlePaddle runtime (install in a separate venv from vllm to avoid conflicts)", "command": "uv pip install paddlepaddle-gpu==3.2.1 --extra-index-url https://www.paddlepaddle.org.cn/packages/stable/cu126/", "brand": "NVIDIA" }, { "note": "PaddleOCR document-parsing helpers — 1.6 requires >= 3.6.0 (pipeline_version v1.6)", "command": "uv pip install -U \"paddleocr[doc-parser]>=3.6.0\"" }, { "note": "Safetensors loader used by the doc-parser path", "command": "uv pip install safetensors" } ] } }, "features": { "text_only": { "description": "Skip loading the vision encoder for text-only workloads — frees VRAM for KV cache. Mutually exclusive with encoder_parallel.", "args": [ "--language-model-only" ] }, "encoder_parallel": { "description": "Run the vision encoder in data-parallel mode — avoids TP comm overhead on the small encoder. Mutually exclusive with text_only.", "args": [ "--mm-encoder-tp-mode", "data" ] } }, "opt_in_features": [ "text_only", "encoder_parallel" ], "variants": { "default": { "precision": "bf16", "vram_minimum_gb": 2, "description": "BF16 weights — small footprint, runs on most GPUs" } }, "kv_offload_support": { "offloading_cpu": "verified", "offloading_fs": "verified" }, "hardware_overrides": { "amd": { "extra_args": [], "extra_env": { "SAFETENSORS_FAST_GPU": "1" } } }, "guide": "## Overview\n\n[PaddleOCR-VL-1.6](https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.6) upgrades\nPaddleOCR-VL-1.5 with a region-aware data optimization framework (targeted enhancement\nof weak regions found in the previous model) and a progressive post-training recipe\nbased on curated data selection and reinforcement learning:\n\n- **96.33% on OmniDocBench v1.6** (new SOTA), plus new records on OmniDocBench v1.5\n and Real5-OmniDocBench (scanning, warping, screen-photography, illumination, skew)\n- Major gains in table, Chinese ancient document, and Chinese rare character recognition\n- Improved seal/stamp recognition, text spotting, and chart recognition\n\nThe architecture is **fully compatible with PaddleOCR-VL-1.5** (same 0.9B NaViT-style\ndynamic-resolution vision encoder + ERNIE-4.5-0.3B LM), so migration is plug-and-play:\nthe vLLM launch command and feature flags are unchanged.\n\n## Prerequisites\n\n- Hardware: 1x GPU (small VRAM footprint)\n- vLLM >= 0.11.1\n\n### Install vLLM\n\n```bash\nuv venv\nsource .venv/bin/activate\nuv pip install -U vllm\n```\n\n## Launch command\n\n```bash\nvllm serve PaddlePaddle/PaddleOCR-VL-1.6 \\\n --trust-remote-code \\\n --max-num-batched-tokens 16384 \\\n --no-enable-prefix-caching \\\n --mm-processor-cache-gb 0\n```\n\nTip: OCR workloads don't benefit much from prefix caching or image reuse, so disable\nthose to avoid hashing/caching overhead.\n\n## Client Usage\n\nTask-specific prompts (same task set as 1.5, including `spotting` and `seal`):\n\n```python\nfrom openai import OpenAI\n\nclient = OpenAI(api_key=\"EMPTY\", base_url=\"http://localhost:8000/v1\", timeout=3600)\n\nTASKS = {\n \"ocr\": \"OCR:\",\n \"table\": \"Table Recognition:\",\n \"formula\": \"Formula Recognition:\",\n \"chart\": \"Chart Recognition:\",\n \"spotting\": \"Spotting:\",\n \"seal\": \"Seal Recognition:\",\n}\n\nresponse = client.chat.completions.create(\n model=\"PaddlePaddle/PaddleOCR-VL-1.6\",\n messages=[{\n \"role\": \"user\",\n \"content\": [\n {\"type\": \"image_url\", \"image_url\": {\"url\": \"https://.../receipt.png\"}},\n {\"type\": \"text\", \"text\": TASKS[\"spotting\"]},\n ],\n }],\n temperature=0.0,\n)\nprint(response.choices[0].message.content)\n```\n\n## Offline Inference with PP-DocLayoutV3\n\nUse separate venvs for `vllm` and `paddlepaddle` to avoid conflicts. The 1.6 pipeline\nrequires `paddleocr[doc-parser] >= 3.6.0` and `pipeline_version=\"v1.6\"`, which\nautomatically uses [PP-DocLayoutV3](https://huggingface.co/PaddlePaddle/PP-DocLayoutV3)\nfor layout detection (PP-DocLayoutV2 belongs to the v1 pipeline). Pass\n`vl_rec_api_model_name=\"PaddlePaddle/PaddleOCR-VL-1.6\"` so the client matches the\nmodel name the vLLM server exposes — otherwise it defaults to\n`PaddleOCR-VL-1.6-0.9B` and fails with \"The model PaddleOCR-VL-1.6-0.9B does not exist.\"\n\n```bash\nuv pip install paddlepaddle-gpu==3.2.1 --extra-index-url https://www.paddlepaddle.org.cn/packages/stable/cu126/\nuv pip install -U \"paddleocr[doc-parser]>=3.6.0\"\nuv pip install safetensors\n```\n\n```python\nfrom paddleocr import PaddleOCRVL\n\npipeline = PaddleOCRVL(\n pipeline_version=\"v1.6\",\n vl_rec_backend=\"vllm-server\",\n vl_rec_server_url=\"http://localhost:8000/v1\",\n vl_rec_api_model_name=\"PaddlePaddle/PaddleOCR-VL-1.6\",\n)\n\noutput = pipeline.predict(\"https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/paddleocr_vl_demo.png\")\nfor i, res in enumerate(output):\n res.save_to_json(save_path=f\"output_{i}.json\")\n res.save_to_markdown(save_path=f\"output_{i}.md\")\n```\n\n## References\n\n- [PaddleOCR-VL-1.6 on Hugging Face](https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.6)\n- [Technical report (arXiv)](https://arxiv.org/pdf/2606.03264)\n- [PaddleOCR GitHub](https://github.com/PaddlePaddle/PaddleOCR)\n- [vLLM support PR](https://github.com/vllm-project/vllm/pull/27758)\n", "source": { "source": "vllm", "repo": "vllm-project/vllm", "ref": "main", "sha": "389ffebe555705c85337025d5df108d0384d365e", "path": "recipes/models/PaddlePaddle/PaddleOCR-VL-1.6.yaml", "edit_url": "https://github.com/vllm-project/vllm/edit/main/recipes/models/PaddlePaddle/PaddleOCR-VL-1.6.yaml", "overlays": [] } }