Add Nanonets-OCR2 documentation to README

- Add Nanonets OCR2 (nanonets-ocr2.py) to available scripts section
- Highlight 3.75B params, enhanced quality, and Qwen2.5-VL base
- Add HF Jobs example for Nanonets OCR2
- Add local usage example
- Position as next-generation quality option

🤖 Generated with Claude Code

README.md

@@ -51,6 +51,19 @@ State-of-the-art document OCR using [nanonets/Nanonets-OCR-s](https://huggingfac
 - 🖼️ **Images** - Captions and descriptions included
 - ☑️ **Forms** - Checkboxes rendered as ☐/☑
 
+### Nanonets OCR2 (`nanonets-ocr2.py`)
+
+Next-generation Nanonets OCR using [nanonets/Nanonets-OCR2-3B](https://huggingface.co/nanonets/Nanonets-OCR2-3B) with improved accuracy:
+
+- 🎯 **Enhanced quality** - 3.75B parameters for superior OCR accuracy
+- 📐 **LaTeX equations** - Mathematical formulas preserved in LaTeX format
+- 📊 **Advanced tables** - Improved HTML table extraction
+- 📝 **Document structure** - Headers, lists, formatting maintained
+- 🖼️ **Smart image captions** - Intelligent descriptions and captions
+- ☑️ **Forms** - Checkboxes rendered as ☐/☑
+- 🌍 **Multilingual** - Enhanced language support
+- 🔧 **Based on Qwen2.5-VL** - Built on state-of-the-art vision-language model
+
 ### SmolDocling (`smoldocling-ocr.py`)
 
 Ultra-compact document understanding using [ds4sd/SmolDocling-256M-preview](https://huggingface.co/ds4sd/SmolDocling-256M-preview) with only 256M parameters:
@@ -155,6 +168,15 @@ hf jobs uv run \
     --max-model-len 16384 \
     --batch-size 128
 
+# Nanonets OCR2 - Next-gen quality with 3B model
+hf jobs uv run \
+    --flavor l4x1 \
+    --secrets HF_TOKEN \
+    https://huggingface.co/datasets/uv-scripts/ocr/raw/main/nanonets-ocr2.py \
+    your-input-dataset \
+    your-output-dataset \
+    --batch-size 16
+
 # NuMarkdown with reasoning traces for complex documents
 hf jobs uv run \
     --flavor l4x4 \
@@ -202,6 +224,9 @@ uv run https://huggingface.co/datasets/uv-scripts/ocr/raw/main/nanonets-ocr.py \
 uv run rolm-ocr.py documents extracted-text
 uv run rolm-ocr.py images texts --shuffle --max-samples 100  # Random sample
 
+# Nanonets OCR2 for highest quality
+uv run nanonets-ocr2.py documents ocr-results
+
 ```
 
 ## 📁 Works With