{"id":18320,"date":"2026-11-18T00:17:09","date_gmt":"2026-11-17T15:17:09","guid":{"rendered":"https:\/\/localmodelwatch.tsuchitsuchi.com\/2026\/11\/18\/open-webui-v0-13-4-released\/"},"modified":"2026-11-18T00:17:09","modified_gmt":"2026-11-17T15:17:09","slug":"open-webui-v0-13-4-released","status":"publish","type":"post","link":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/18\/open-webui-v0-13-4-released\/","title":{"rendered":"Open WebUI v0.13.4 Released: Tether and Coding Agents"},"content":{"rendered":"<p><!-- lmw:facts --><\/p>\n<h2>At a Glance<\/h2>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Item<\/th>\n<th>Value<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Repository<\/td>\n<td><a href=\"https:\/\/github.com\/open-webui\/open-webui\">open-webui\/open-webui<\/a><\/td>\n<\/tr>\n<tr>\n<td>Version<\/td>\n<td><a href=\"https:\/\/github.com\/open-webui\/open-webui\/releases\/tag\/v0.13.4\">v0.13.4<\/a><\/td>\n<\/tr>\n<tr>\n<td>Published<\/td>\n<td>2026-11-17<\/td>\n<\/tr>\n<tr>\n<td>License<\/td>\n<td>NOASSERTION<\/td>\n<\/tr>\n<tr>\n<td>Source type<\/td>\n<td>Primary source (the publisher itself)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Values determined by this site&#8217;s code when the information was collected. Dates are JST.<\/em><\/p>\n<p><!-- \/lmw:facts --><\/p>\n<h2>Overview<\/h2>\n<p>Open WebUI is a web interface (primarily built with Svelte) that connects to Ollama and OpenAI-compatible APIs. This v0.13.4 release was published on 2026-11-17.<\/p>\n<p>The most impactful change is the introduction of Tether, which replaces Direct Connections. Users currently relying on Direct Connections will need to prepare for migration before updating. Everyone else can continue using the software as before unless they choose to use Tether.<\/p>\n<h2>Breaking Changes and Deprecations<\/h2>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Old<\/th>\n<th>New<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Direct Connections<\/td>\n<td>Tether (Tether 0.1.14 or later; MCP servers require 0.1.16 or later)<\/td>\n<\/tr>\n<tr>\n<td>Model information received by Functions \/ Tools includes profile images<\/td>\n<td>Models are passed without profile images<\/td>\n<\/tr>\n<tr>\n<td>Chats from older versions lack terminal storage<\/td>\n<td>Terminals and working folders are saved per chat (older chats open without a terminal, excluding coding agent chats)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>Users of Direct Connections must switch to Tether. The release notes do not provide a migration guide from Direct Connections. To use tool servers, MCP servers, and terminals via Tether, <code>ENABLE_TOOL_SERVERS<\/code> must be enabled. If a Tether model ID matches an existing model on the Open WebUI side, the Tether model will be hidden. This also affects code in Functions and Tools that reads model profile images.<\/p>\n<h2>Key Changes<\/h2>\n<h3>Connecting Local Computers with Tether<\/h3>\n<p>Models from Ollama, LM Studio, and OpenAI-compatible servers added to the Tether app appear as personal models in the connected Open WebUI model selection menu. OpenAPI\/MCP tool servers and Open Terminal servers work similarly. Addresses, commands, and keys never leave your local PC.<\/p>\n<p>Starting with Tether 0.1.13, terminals can be sandboxed. This applies to macOS and supported Windows systems, and is enabled via Tether under Settings &gt; Open Terminal &gt; Sandbox. If sandboxing cannot be enforced, the terminal remains stopped.<\/p>\n<p>Local MCP servers running via stdio can also be used without proxies like mcpo starting from Tether 0.1.16. However, local commands run with your account permissions and reside outside the Tether sandbox.<\/p>\n<h3>User-Specific Terminals and Coding Agents<\/h3>\n<p>When connected to Terminals 0.4.0 or later, users with write permissions to the connection can create sandboxed terminals from the terminal menu in the message input field. You can also choose the sharing scope (one for all users, or per user, chat, or model). Starting and stopping requires Terminals 0.4.1 or later. Open Terminal 0.17.7 or later inside the terminal image is required for startup directories and setup scripts.<\/p>\n<p>Coding agents can be used in chat just like regular models. Hermes, Antigravity, GitHub Copilot, Goose, and Qwen Code have been added, alongside ACP agents added via <code>OPEN_TERMINAL_ACP_AGENTS<\/code>. New agents require Open Terminal 0.17.19 or later, as well as installation and sign-in on the terminal side.<\/p>\n<h3>Image and Video Generation<\/h3>\n<p>Administrators can add multiple image connections under Admin Settings &gt; Images. Supported backends include OpenAI-compatible servers such as LiteLLM, Gemini, ComfyUI, and AUTOMATIC1111. Built-in video generation tools can be enabled under Admin Settings &gt; Videos by adding OpenAI, LiteLLM, or OpenRouter connections. The Video Generation permission is disabled by default.<\/p>\n<h3>Folders and UI<\/h3>\n<p>The folder page now features Chat List, Board view, Status, Automations, and Knowledge sections. Notes, channels, and workspaces can be opened in tabs. Side chats function as the same chat as the main view. The Visualizations built-in tool is disabled by default.<\/p>\n<h3>Performance Improvements<\/h3>\n<ul>\n<li><strong>Responses with heavy code blocks<\/strong>: Freezing during streaming has been reduced (PR #34495). Displaying 300 code blocks at once dropped from 143 seconds to 9 seconds.<\/li>\n<li><strong>Large PDF previews<\/strong>: Opening a 1,500-page text PDF at the 1,200th page improved from 35.7 seconds and 7.5 GB to 1.3 seconds and 1.0 GB (PR #34602). Browser Ctrl+F now targets only pages near the screen.<\/li>\n<li><strong>Model lists when using Redis<\/strong>: Profile images are no longer stored in Redis (PR #34595). For 4 images (total storage size of 20 MB), list loading time decreased from 64 ms to 45 ms.<\/li>\n<li><strong>Profile image and knowledge exports<\/strong>: Loading a single 2 MB model image improved from 207 ms to 15 ms (PR #34598).<\/li>\n<\/ul>\n<p>The documentation notes that the end of the release notes was cut off, so additional changes may be included.<\/p>\n<h2>Hardware Requirements<\/h2>\n<p>Here is a summary of newly supported features from a user&#8217;s perspective. The documentation does not mention support for new GPUs, accelerators, or quantization formats.<\/p>\n<ul>\n<li><strong>Coding Agents<\/strong>: Hermes, Antigravity, GitHub Copilot, Goose, and Qwen Code can now be used as chat models. Any custom ACP agent added via <code>OPEN_TERMINAL_ACP_AGENTS<\/code> will also work, provided they are installed and signed in on the terminal side.<\/li>\n<li><strong>Video Generation Models<\/strong>: Videos can be generated via OpenAI, LiteLLM, and OpenRouter connections. The release notes list Veo and Kling as models accessed through OpenRouter. These are cloud-based models and not a feature for running local open-weight models.<\/li>\n<li><strong>Image Generation Backends<\/strong>: Multiple OpenAI-compatible servers such as LiteLLM, Gemini, ComfyUI, and AUTOMATIC1111 can be connected simultaneously. If no model ID is entered, image models from the LiteLLM server automatically appear in the list. Users connecting local ComfyUI or AUTOMATIC1111 instances can configure separate default models for generation and editing.<\/li>\n<li><strong>OpenAI-Compatible Vector Stores<\/strong>: Vector stores on OpenAI, Azure OpenAI, and LiteLLM can be connected as external knowledge sources. Because search is handled server-side, corresponding embedding models are not required.<\/li>\n<li><strong>Tether Sandbox<\/strong>: Available on macOS and supported Windows devices. According to the release notes, Docker, virtual machines, and administrator privileges are not required. Linux support is not mentioned in the documentation.<\/li>\n<\/ul>\n<h2>How to Get It<\/h2>\n<p>Installation steps described in the README are as follows. The release notes do not provide separate update instructions specifically for v0.13.4.<\/p>\n<h3>For Python<\/h3>\n<p>If using uv:<\/p>\n<pre><code class=\"language-bash\">DATA_DIR=~\/.open-webui uvx --python 3.12 open-webui@latest serve\n<\/code><\/pre>\n<p>If using pip (Python 3.12):<\/p>\n<pre><code class=\"language-bash\">pip install open-webui\nopen-webui serve\n<\/code><\/pre>\n<p>After starting either command, open http:\/\/localhost:8080.<\/p>\n<h3>For Docker<\/h3>\n<p>The README command is:<\/p>\n<pre><code class=\"language-bash\">docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:\/app\/backend\/data --name open-webui --restart always ghcr.io\/open-webui\/open-webui:main\n<\/code><\/pre>\n<p>After starting, open http:\/\/localhost:3000. Choose the image tag based on your use case:<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Tag<\/th>\n<th>Description<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>:main<\/code><\/td>\n<td>Standard image including local AI dependencies<\/td>\n<\/tr>\n<tr>\n<td><code>:slim<\/code><\/td>\n<td>Lightweight version using external services for embeddings, document extraction, and audio<\/td>\n<\/tr>\n<tr>\n<td><code>:cuda<\/code><\/td>\n<td>CUDA-enabled (requires <code>--gpus all<\/code> and the Nvidia Container Toolkit)<\/td>\n<\/tr>\n<tr>\n<td><code>:ollama<\/code><\/td>\n<td>Bundled with Ollama<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<h3>Checks After Updating<\/h3>\n<ul>\n<li>If using Tether, make sure it is updated to Tether 0.1.14 or later (MCP servers require 0.1.16 or later).<\/li>\n<li>New terminal-related features may require Open Terminal 0.17.19 or later, or Terminals 0.4.0 or later, as specified in the relevant sections above.<\/li>\n<li>For detailed Docker configuration, refer to the <a href=\"https:\/\/docs.openwebui.com\/getting-started\/quick-start\/\">Docker installation guide<\/a>.<\/li>\n<\/ul>\n<p><!-- lmw:releases --><\/p>\n<h2>Releases Since Our Last Article<\/h2>\n<p><em>Compiled by Local Model Watch from the project&#8217;s GitHub releases: the versions between this release and the last one we covered, which did not get separate articles.<\/em> <em>Full history: <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-open-webui-en\/\">release tracker<\/a>.<\/em><\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Version<\/th>\n<th>Released<\/th>\n<th>Release notes<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>v0.13.3<\/td>\n<td>2026-11-09<\/td>\n<td><a href=\"https:\/\/github.com\/open-webui\/open-webui\/releases\/tag\/v0.13.3\">GitHub<\/a><\/td>\n<\/tr>\n<tr>\n<td>v0.13.2<\/td>\n<td>2026-11-08<\/td>\n<td><a href=\"https:\/\/github.com\/open-webui\/open-webui\/releases\/tag\/v0.13.2\">GitHub<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><!-- \/lmw:releases --><\/p>\n<p><!-- lmw:related --><\/p>\n<h2>Related Articles<\/h2>\n<ul>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/open-webui-v0131-release\/\">Open WebUI v0.13.1 Released with Security Fixes and Tether<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/open-webui-v0-13-0\/\">Open WebUI v0.13.0 Released: Tabbed Workspaces &amp; Security Fixes<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/open-webui-v0-12-4-released\/\">Open WebUI v0.12.4 Released with Context Compaction &amp; On-Demand Skills<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/open-webui-v0-12-3-released\/\">Open WebUI v0.12.3 Released with E2EE Chats and Redis Streaming<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:related --><\/p>\n<p><!-- lmw:next-steps --><\/p>\n<h2>What to Read Next<\/h2>\n<ul>\n<li><strong>Follow this tool<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-open-webui-en\/\">Open WebUI overview and release history (182 releases tracked)<\/a><\/li>\n<li><strong>Other uis and desktop apps<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-jan-en\/\">Jan<\/a> \/ <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-text-generation-webui-en\/\">TextGen<\/a><\/li>\n<li><strong>Engines mentioned in this article<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-ollama-en\/\">Ollama<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:next-steps --><\/p>\n<h2>Sources<\/h2>\n<ul>\n<li><a href=\"https:\/\/github.com\/open-webui\/open-webui\/releases\/tag\/v0.13.4\">open-webui\/open-webui v0.13.4 Release Notes<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/open-webui\/open-webui\/pull\/34495\">PR #34495 Fix freezing on responses with heavy code blocks<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/open-webui\/open-webui\/pull\/34602\">PR #34602 Speed up large PDF previews<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/open-webui\/open-webui\/pull\/34595\">PR #34595 Speed up model list loading when using Redis<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/open-webui\/open-webui\/pull\/34598\">PR #34598 Speed up profile image and knowledge exports<\/a><\/li>\n<li><a href=\"https:\/\/docs.openwebui.com\/getting-started\/quick-start\/\">Docker installation guide<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Open WebUI v0.13.4 is released with Tether replacing Direct Connections, new coding agent integrations, performance boosts, and more.<\/p>\n","protected":false},"author":1,"featured_media":18319,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[105],"tags":[1475,1655,1093,1917,4766,1547],"class_list":["post-18320","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-engines-and-tools","tag-docker-en","tag-mcp-en","tag-ollama-en","tag-open-webui-en","tag-tether-en","tag-verified"],"lang":"en","translations":{"en":18320,"ja":18318},"_links":{"self":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/18320","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/comments?post=18320"}],"version-history":[{"count":0,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/18320\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media\/18319"}],"wp:attachment":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media?parent=18320"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/categories?post=18320"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/tags?post=18320"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}