Commit Graph
18920 Commits
Author SHA1 Message Date
Timothy Jaeryang Baek a3a2e42ee0 refac 2026-10-05 12:53:56 +04:00
Timothy Jaeryang Baek 67e725c198 refac 2026-10-05 12:49:22 +04:00
Timothy Jaeryang Baek 5bb1c47080 refac 2026-10-05 12:49:11 +04:00
Timothy Jaeryang Baek 24e30d1cbd refac 2026-10-05 12:17:25 +04:00
Timothy Jaeryang Baek 38cda57c86 refac 2026-10-05 12:03:32 +04:00
Timothy Jaeryang Baek 250f63175e refac 2026-10-05 12:03:25 +04:00
Timothy Jaeryang Baek 425da8b6cc refac 2026-10-05 12:03:19 +04:00
ce94c41d81 fix(chat): support unarchive toggle and dynamic toast in chat header menu (#31473) (#31857)
- Pass `archived` state from active chat model to Navbar context menu
- Render 'Unarchive' instead of 'Archive' when `chat?.archived` is true in Navbar/Menu.svelte
- Dynamically display 'Chat unarchived.' toast upon unarchiving in archiveChatHandler
- Closes #31473

Co-authored-by: PINYO PATTANAWASANPORN <pattanawasanporn@gmail.com>
2026-10-05 11:20:02 +04:00
Classic298andGitHub 88720c6928 fix: tool calls get mixed up between rounds when the model reuses tool call ids (#31887)
Some providers, like Kimi K3 on OpenRouter, number their tool calls from zero again each time the model calls tools within the same reply. A later batch of calls then overwrote the earlier calls with the same id, so the saved chat showed the earlier calls with the later arguments, and the model was sent its earlier results next to the wrong arguments. A new tool call whose id is already used in the same reply now gets a fresh id, so every call keeps its own arguments and result, and providers that send unique ids are untouched. Tested before and after against a mock provider that reuses ids on every round.

Fixes #28305
2026-10-05 11:19:45 +04:00
Timothy Jaeryang Baek 53f47b3dc7 refac 2026-10-05 11:19:23 +04:00
Classic298andGitHub 878c90eaaa fix: terminal and file browser jump back to the top folder after every command (#31878)
The terminal runs commands in whatever folder the file browser panel is showing. When the file browser is limited to a top folder (OPEN_TERMINAL_FILE_BROWSER_ROOT, set to the home folder by default), every time the agent ran a command in the terminal the panel jumped back to that top folder and closed the open file, so the next command ran there too. Now the panel stays in the folder you picked and just reloads its contents after each command, keeping the open file in view.

Fixes #30051
2026-10-05 11:15:42 +04:00
Classic298andGitHub 273180a0d7 fix: Read Aloud and auto-playback ignore the Speech Playback Speed setting (#31881)
Read Aloud and response auto-playback always played at normal speed, regardless of the Speech Playback Speed setting. The speech is played one sentence at a time, and the browser put the speed back to normal every time a new sentence started. The chosen speed now stays in effect for every sentence. Voice mode already respected the setting.

Fixes #31870
2026-10-05 11:15:06 +04:00
G30andGitHub 7dd9fc7076 fix: stop the rest of the reply when interrupting Voice mode (#31876) (#31877) 2026-10-05 11:13:38 +04:00
Timothy Jaeryang Baek 4e1c52696f refac 2026-10-05 11:11:42 +04:00
Timothy Jaeryang Baek fb741ebcd2 refac 2026-10-05 10:56:06 +04:00
Timothy Jaeryang Baek cf5755f949 refac 2026-10-05 10:48:41 +04:00
Timothy Jaeryang Baek 7e7aeb48cd refac 2026-10-05 10:46:15 +04:00
Timothy Jaeryang Baek d8659c237c refac 2026-10-05 10:46:08 +04:00
Timothy Jaeryang Baek dd576bade3 refac 2026-10-05 10:40:09 +04:00
G30andGitHub d4a2d11ecf fix: hide Share in the function menu when Community Sharing is off (#31820) 2026-10-05 10:36:51 +04:00
Timothy Jaeryang Baek 079647f142 refac 2026-10-05 10:27:07 +04:00
Timothy Jaeryang Baek e608e89b9e refac 2026-10-05 10:03:01 +04:00
Timothy Jaeryang Baek 77e6bc2893 refac 2026-10-05 09:56:10 +04:00
Timothy Jaeryang Baek 7977cdc343 refac 2026-10-05 09:46:43 +04:00
Timothy Jaeryang Baek 51f0e01258 refac 2026-10-05 09:24:43 +04:00
Timothy Jaeryang Baek fd0749fba1 refac 2026-10-05 06:58:07 +04:00
Timothy Jaeryang Baek 89ed348afd refac 2026-10-05 06:56:31 +04:00
Classic298andGitHub 983fa6c974 fix: mixed Arabic and English lines in a message all take the first line's direction (#31868)
When a message mixed right-to-left and left-to-right lines, the chat input and the sent message gave every line the direction of the first one, so an English line under an Arabic one showed up as "?Hello, how are you" with the question mark at the start and sat on the wrong side. Each line now picks its own direction and alignment, both while typing and in the sent message. With Chat Direction set to LTR or RTL instead of Auto, lines in the input box now also align by their own language, the way sent messages already did.

Fixes #31827
2026-10-05 06:49:02 +04:00
Classic298andGitHub f25708484b fix: Daily Messages chart in Analytics fails to load on All time when a message has no timestamp (#31888)
With All time selected, the Daily Messages chart in Admin Panel > Analytics failed to load (the daily analytics request returned a 500) as soon as one message had been saved without a timestamp, for example a chat created through the API with a null timestamp. A date range still worked because those messages were left out of it. Such messages now get the time they were saved, and ones already stored without a timestamp are counted on today's date instead of breaking the chart. Messages that came after the missing timestamp in the same chat were also left out of analytics before and are now counted.

Fixes #27316
2026-10-05 06:46:49 +04:00
Classic298andGitHub c5dac0070d fix: pasted emails are sent with the same text repeated many times (#31890)
With Rich Text Input for Chat on, pasting an HTML email or newsletter into the chat input looked fine, but the sent message repeated the same text over and over as Markdown table rows. These emails are laid out with tables nested inside other tables, and the content of each inner table was counted again for every table around it. Pasted content is now sent once, and regular tables are unaffected.

Fixes #24657
2026-10-05 06:46:33 +04:00
Classic298andGitHub 808b90c031 fix: line breaks in chat messages disappear after reloading the page (#31869)
Lines separated by a single newline show on separate lines while you chat, but after reloading the page or opening a chat by its URL they run together into one line, and on shared chat links they always did. The single-newline line break setting was only switched on once the chat input had loaded, so any message shown before that rendered without it. It is now on from the start, so messages look the same after a reload, from a URL and on shared links as they do while chatting.
2026-10-05 06:45:51 +04:00
Classic298andGitHub 1e834c09e2 fix: Web API dictation stops while you are still talking (#31891)
With the Web API speech-to-text engine, the two-second silence timer only restarted when the browser finished a phrase. If you paused briefly and then kept talking for more than two seconds, dictation stopped mid-sentence. The timer now restarts whenever the browser hears speech, so dictation only stops after two seconds of silence.

Fixes #19727
2026-10-05 06:45:11 +04:00
de9234a2b8 fix: keep edited replies searchable by their new text (#31844)
Saving an edited reply stored the new structured output but set the message's plain content to an empty string. Chat search only matches message content, so the chat could no longer be found by the edited text, nor by the text it replaced.

editMessage now fills content from the edited output with the existing getOutputText, the same text the reply is shown with. Saving as a copy does the same. Output still takes precedence for rendering and for the next request to the model, so only the searchable copy changes.

Fixes #31471

Co-authored-by: RaulMermans <214883465+RaulMermans@users.noreply.github.com>
2026-10-05 06:43:49 +04:00
G30andGitHub 41ca360ccf fix: stop Ctrl+Shift+Enter from sending the message when Ctrl+Enter to Send is on (#31864) 2026-10-05 06:43:14 +04:00
G30andGitHub 1b2ceedd61 fix: create the model when a GGUF is uploaded to Ollama, in File Mode and URL Mode (#31862) 2026-10-05 06:43:06 +04:00
G30andGitHub 8c34af3033 fix: keep private arena models with no access grants visible to admins without the admin bypass (#31858) 2026-10-05 06:42:55 +04:00
G30andGitHub 2e43d66985 fix: report a web search where no page loaded instead of blaming the embedding settings (#31859) 2026-10-05 06:42:48 +04:00
Classic298andGitHub 8e03389de1 fix: inline math touching an em dash or en dash is not rendered (#31901)
In chat replies, inline math written right next to an em dash or en dash, with either $...$ or \(...\), was left unrendered, as in "constants—\(e\), \(i\) and \(0\)—in one equation" where the 0 showed up as plain "(0)". Both dashes are now allowed directly before or after a formula, like spaces, commas and brackets already are, so the math renders whenever a dash touches it.
2026-10-05 06:41:57 +04:00
Classic298andGitHub 95dd3321af fix: images stored inside a chat trigger needless context compaction (#31915)
Images stored inside the chat record itself (chats from older versions, images that could not be saved as files, temporary chats) had their encoded data counted as text when estimating how full the context is, so a single 3 MB image counted as about a million tokens. That pushed the chat over the compaction threshold, summarizing older messages while the chat was well under the limit, and made the context usage indicator jump. The encoded image data is now left out of the token estimate, both for compaction and for the indicator.

Fixes #31913
2026-10-05 06:41:50 +04:00
f9dbe200af i18n: fill missing ja-JP translations (#31911)
Translate empty and missing Japanese UI strings, including plural forms that Japanese does not distinguish, and correct a few existing labels.

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
2026-10-05 06:40:39 +04:00
Classic298andGitHub 743a46bdcd fix: chat stays broken after a reply is stopped or fails before any text (#31892)
When a reply was stopped or failed before the model sent any text, the chat kept an empty reply, and every later message in that chat sent it back to the model as part of the conversation. Providers that reject empty assistant messages then refused every new request, so the chat stayed unusable unless the empty reply was deleted from the database by hand. Empty replies are now skipped when the conversation is sent to the model, as replies that ended in an error already were. Chats already stuck like this work again on their next message, since nothing stored has to change.

Fixes #25083
2026-10-03 17:06:03 +04:00
Classic298andGitHub 914ff1f116 fix: large pastes into a note are lost after a reload (#31893)
Pasting a large block of text (around 150 KB or more) into a note made the note's live connection drop and reconnect, and the text was never saved, so a reload showed the note without it. Each edit message carries the whole note in several formats, which pushed it past the server's 1 MB cap on a single message. That cap is now 16 MiB, the same size the server already accepts for a websocket message, so notes with up to about 2 MB of text save again.

Fixes #26140
2026-10-03 17:05:00 +04:00
Classic298andGitHub cd64930c05 fix: MCP OAuth sign-in fails with a state error when the server asks for many scopes (#31894)
Connecting an MCP tool server over OAuth 2.1 failed after signing in at the provider with an "invalid or expired state" error whenever the server advertises a long list of scopes, such as the Google Workspace MCP server with its 42 Google scopes. Open WebUI saved the full authorization link it sends to the provider in the session cookie, which pushed the cookie past the 4096-byte browser limit, so the browser dropped it and Open WebUI could not recognise the user when the provider sent them back. The link is no longer saved there, so the cookie stays at a few hundred bytes even with long scope lists.

Fixes #26382
2026-10-03 17:04:50 +04:00
Classic298andGitHub 394295ec76 fix: older responses keep a loading cursor forever after opening a chat (#31895)
Opening a chat only cleared the loading cursor on the newest response that never finished, so an older one (from a cancelled generation, or a turn an earlier Open WebUI version left unfinished) kept spinning and never showed its copy and regenerate buttons. Now, when no generation is running in the chat, every unfinished response shows as finished on load, except one waiting for the user to answer a tool's question. While a generation is running nothing changes, because another model's response to the same message may still be streaming in.

Fixes #14806
2026-10-03 17:04:29 +04:00
G30andGitHub 7e6c8d7b6f fix: hide the user menu pin buttons when the window loses focus (#31841) 2026-10-02 08:50:54 +04:00
Classic298andGitHub 8b4d3ac527 fix: chat freezes when reopened after tool calls with some Responses API providers (#31838)
Some providers, such as OpenVINO Model Server, number their output items from zero again in every response. When a model wrote text, called a tool and then answered, the saved message held two items with the same id, and opening that chat later locked up the browser tab. Repeated ids are now renamed when the chat is shown, so these chats open normally, including ones saved before this fix. Providers that already send unique ids render exactly as before.

Fixes #31837
2026-10-02 08:50:44 +04:00
G30andGitHub ecaa1a1bbc fix: update the user menu pin button icon and tooltip when it is clicked (#31843) 2026-10-02 08:49:11 +04:00
G30andGitHub 571dafae36 fix: show unsaved model editor changes in the JSON Preview (#31851) 2026-10-02 08:32:55 +04:00
Timothy Jaeryang Baek fe2c694915 refac 2026-10-02 08:24:53 +04:00
G30andGitHub 6eb0a0b77e fix: stop deleting the model when an Ollama download is cancelled (#31824) 2026-10-01 12:26:13 -05:00