[PR #14787] [MERGED] feat: Customize Docling's "Describe Pictures" feature #127251

New Issue

GiteaMirror · 2026-05-21T09:00:16-05:00

GiteaMirror commented

2026-05-21 09:00:16 -05:00

📋 Pull Request Information

Original PR: https://github.com/open-webui/open-webui/pull/14787
Author: @vaclcer
Created: 6/8/2025
Status: ✅ Merged
Merged: 6/8/2025
Merged by: @tjbck

Base: dev ← Head: vaclavs-custom-docling

📝 Commits (1)

99f0556 Add configuration options for picture description modes and update related components

📊 Changes

5 files changed (+311 additions, -7 deletions)

View changed files

📝 backend/open_webui/config.py (+48 -0)
📝 backend/open_webui/main.py (+14 -0)
📝 backend/open_webui/retrieval/loaders/main.py (+51 -7)
📝 backend/open_webui/routers/retrieval.py (+64 -0)
📝 src/lib/components/admin/Settings/Documents.svelte (+134 -0)

📄 Description

Description

This PR adds a customization to the Docling's content extraction engine. It follows recent Docling's development described at https://github.com/docling-project/docling-serve/blob/main/docs/usage.md#picture-description-options.

With this PR, user can choose to use a locally hosted VLM (in Docling env) to describe pictures in the document or use external OpenAI-like API to do so.

Additional Information

Please advice if this is wanted or not, especially UI styling etc.

Screenshots or Videos

Contributor License Agreement

By submitting this pull request, I confirm that I have read and fully agree to the Contributor License Agreement (CLA), and I am providing my contributions under its terms.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/open-webui/open-webui/pull/14787 **Author:** [@vaclcer](https://github.com/vaclcer) **Created:** 6/8/2025 **Status:** ✅ Merged **Merged:** 6/8/2025 **Merged by:** [@tjbck](https://github.com/tjbck) **Base:** `dev` ← **Head:** `vaclavs-custom-docling` --- ### 📝 Commits (1) - [`99f0556`](https://github.com/open-webui/open-webui/commit/99f05561f85c7951f26585338ceaa58b5b05c719) Add configuration options for picture description modes and update related components ### 📊 Changes **5 files changed** (+311 additions, -7 deletions) <details> <summary>View changed files</summary> 📝 `backend/open_webui/config.py` (+48 -0) 📝 `backend/open_webui/main.py` (+14 -0) 📝 `backend/open_webui/retrieval/loaders/main.py` (+51 -7) 📝 `backend/open_webui/routers/retrieval.py` (+64 -0) 📝 `src/lib/components/admin/Settings/Documents.svelte` (+134 -0) </details> ### 📄 Description ### Description This PR adds a customization to the Docling's content extraction engine. It follows recent Docling's development described at https://github.com/docling-project/docling-serve/blob/main/docs/usage.md#picture-description-options. With this PR, user can choose to use a locally hosted VLM (in Docling env) to describe pictures in the document or use external OpenAI-like API to do so. ### Additional Information Please advice if this is wanted or not, especially UI styling etc. ### Screenshots or Videos ![describe1](https://github.com/user-attachments/assets/f3b5741c-fa87-4dc2-ab11-5d53fe90f0bb) ![describe2](https://github.com/user-attachments/assets/5ebcff21-213e-4166-ab62-5e4762effb1a) ### Contributor License Agreement By submitting this pull request, I confirm that I have read and fully agree to the [Contributor License Agreement (CLA)](/CONTRIBUTOR_LICENSE_AGREEMENT), and I am providing my contributions under its terms. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

GiteaMirror added the pull-request label 2026-05-21 09:00:16 -05:00

GiteaMirror closed this issue

2026-05-21 09:00:18 -05:00

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/open-webui#127251