docs: OCR v2 polish on multimodal-extraction (Helm/local HF)

docs/docs/extraction/multimodal-extraction.md

@@ -60,7 +60,7 @@ Charts and infographic regions are classified with other page layout elements (t
 
     For chart and infographic detection and modality-specific retrieval, use the default **pdfium** layout path (page-elements detection and OCR), not `extract_method="nemotron_parse"`.
 
-For how chart-labeled PDF regions interact with captioning, refer to [Image captioning](#image-captioning).
+Chart-labeled PDF regions are **not** routed through the Omni caption stage; they remain on the layout-and-OCR path. For scope and validation guidance, refer to [Image captioning](#image-captioning).
 
 For natural-language infographic descriptions, optionally enable [image captioning](#image-captioning) and set `caption_infographics=True` when you need VLM captions on infographic regions.
 
@@ -74,7 +74,13 @@ For natural-language infographic descriptions, optionally enable [image captioni
 
 Scanned PDFs and image-only pages rely on OCR and hybrid paths that combine native text extraction with OCR when needed. For extract methods such as `ocr` and `pdfium_hybrid`, refer to the [Python API reference](nemo-retriever-api-reference.md).
 
-When you run extraction locally with Hugging Face weights, the default OCR engine is **Nemotron OCR v2**, which operates in **multilingual** mode by default. For CLI flags and API parameters, refer to [Nemotron OCR v2 — language mode](https://github.com/NVIDIA/NeMo-Retriever/blob/main/nemo_retriever/docs/cli/README.md#nemotron-ocr-v2-language-mode). For Kubernetes deployment, refer to [OCR NIM configuration](https://github.com/NVIDIA/NeMo-Retriever/blob/main/nemo_retriever/helm/README.md#ocr-nim-configuration) in the Helm chart README.
+OCR artifacts depend on how you deploy.
+
+**Helm / NIM:** The production chart deploys **Nemotron OCR v2**. For image defaults and upgrade notes, refer to [OCR NIM configuration](https://github.com/NVIDIA/NeMo-Retriever/blob/main/nemo_retriever/helm/README.md#ocr-nim-configuration) in the Helm chart README.
+
+**Local Hugging Face inference:** The default engine is **Nemotron OCR v2**, which operates in **multilingual** mode by default. For CLI flags and API parameters, refer to [Nemotron OCR v2 — language mode](https://github.com/NVIDIA/NeMo-Retriever/blob/main/nemo_retriever/docs/cli/README.md#nemotron-ocr-v2-language-mode).
+
+For Kubernetes defaults and the Helm-vs-local split, refer to [OCR artifacts (Helm vs local Hugging Face)](prerequisites-support-matrix.md#nemotron-ocr-v2-language-mode) in the support matrix.
 
 **Related**
 

docs/docs/extraction/releasenotes.md

@@ -36,9 +36,9 @@ Highlights for the 26.05 release include:
 
 ### Models, OCR, and captioning
 
-- Nemotron OCR v2 is the default OCR engine for HuggingFace, with CLI language selectors and unified OCR actors. For Helm NIM deployments, Nemotron OCR v1 is the default.  
-- Nemotron Parse is available as an alternate PDF extraction method (v1.2 HTTP interface; optional Helm NIM; local inference via vLLM where configured)  
-- VLM image captioning via vLLM (including Omni caption model profiles) addresses the capability deferred in 26.03  
+- Nemotron OCR v2 is the default OCR engine for Hugging Face and Helm NIM deployments, with CLI language selectors and unified OCR actors.  
+- Nemotron Parse is available as an alternate PDF extraction method (v1.2 HTTP interface; optional Helm NIM; local inference with vLLM where configured)  
+- VLM image captioning with vLLM (including Omni caption model profiles) addresses the capability deferred in 26.03  
 - vLLM-backed text and vision-language embedders, multimodal VL reranker, and torch 2.11 for local GPU installs  
 
 ### Multimodal extraction
@@ -48,7 +48,7 @@ Highlights for the 26.05 release include:
 
 ### Retrieval and RAG
 
-- Live RAG SDK with `Retriever.retrieve()`,  reference answer generation `Retriever.answer()`, and optional batch operator graphs via LiteLLM (`[llm]` extra)  
+- Live RAG SDK with `Retriever.retrieve()`, reference answer generation `Retriever.answer()`, and optional batch operator graphs through LiteLLM (`[llm]` extra)  
 
 ### Vector database