docs: sync with latest GitHub (OpenBMB links, examples, KIE, seal/chart prompts)

.gitattributes

@@ -42,3 +42,4 @@ assets/minicpm-english-examples-3-column.gif filter=lfs diff=lfs merge=lfs -text
 assets/minicpm-ocr-english-cases.png filter=lfs diff=lfs merge=lfs -text
 assets/ocr-benchmarks.png filter=lfs diff=lfs merge=lfs -text
 assets/throughput-comparison.png filter=lfs diff=lfs merge=lfs -text
+assets/minicpm-ocr-chinese-cases.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -84,23 +84,27 @@ Real throughput also depends on page content, backend, concurrency, and hardware
 
 ## Examples
 
-The animations show the original page, the detected layout boxes, and the rendered Markdown.
-Click an image to open it at full resolution.
+Left to right: original document, detected layout, and rendered Markdown.
+
+**English documents**
+
+<p align="center"><img src="./assets/minicpm-english-examples-3-column.gif" alt="English document animation: original, detected layout, and rendered Markdown." width="100%"></p>
 
 **Chinese documents**
 
-<p align="center"><img src="./assets/minicpm-chinese-examples-3-column.gif" alt="Chinese document parsing: original page, detected layout boxes, and rendered Markdown." width="100%"></p>
+<p align="center"><img src="./assets/minicpm-chinese-examples-3-column.gif" alt="Chinese document animation: original, detected layout, and rendered Markdown." width="100%"></p>
 
-**English documents**
+### More examples
 
-<p align="center"><img src="./assets/minicpm-english-examples-3-column.gif" alt="English document parsing: original page, detected layout boxes, and rendered Markdown." width="100%"></p>
+Original documents and parsed results, side by side.
 
-### Static examples
+**English documents**
+
+<p align="center"><img src="./assets/minicpm-ocr-english-cases.png" alt="English examples comparing inputs and results: handwriting, historical and scanned documents, newspapers, financial statements, and key information extraction." width="100%"></p>
 
-Handwritten notes, historical documents, scans, newspapers, financial statements, and key
-information extraction — originals next to parse results.
+**Chinese documents**
 
-<p align="center"><img src="./assets/minicpm-ocr-english-cases.png" alt="MiniCPM-OCR 1.0 English examples: original inputs and parsing results for handwritten notes, historical documents, scanned documents, newspapers, financial statements, and key information extraction." width="100%"></p>
+<p align="center"><img src="./assets/minicpm-ocr-chinese-cases.png" alt="Chinese examples comparing inputs and results: handwritten notes, complex formulas, newspapers, financial statements, stamped documents, illustrations, and key information extraction." width="100%"></p>
 
 ## Model architecture
 
@@ -165,9 +169,27 @@ The prompt selects the task and must match the region you cropped:
 | Table | `Table Recognition:` | HTML (via SDK post-processing) |
 | Formula | `Formula Recognition:` | LaTeX |
 | Inline formula | `Text Recognition:` | LaTeX |
+| Seal | `Seal Recognition:` | seal text |
+| Chart | `Chart to Table:` | table |
 
 If `finish_reason` is `length`, the output was truncated — raise the token budget.
 
+### Key information extraction (KIE)
+
+MiniCPM-OCR 1.0 can pull specified fields straight out of an image — for example the store name
+and amount on a receipt. Send the image with a field-extraction prompt and ask for a JSON object;
+you do not need to produce page Markdown first.
+
+Use the [image recognition API example](https://github.com/OpenBMB/MiniCPM-OCR/blob/main/docs/vllm_deploy_en.md#22-image-recognition) and
+replace `Text Recognition:` with:
+
+```text
+Extract the "store_name", "amount". Format the output as a JSON object with keys: "store_name", "amount". If an item is not in the image, use "###".
+```
+
+Replace both field lists with your own field names (Chinese or English). The response is
+model-generated text: parse the JSON and check fields and values before using it downstream.
+
 ### Agent skill
 
 Have a coding agent follow the skill to set up the environment, download the model, and parse