0.6.0

🚀 Features

• ✨enhancement: Model sources should not be hard coded from frontend @louis-menlo (#5270)
• ✨enhancement: add setting chat width container @urmauur (#5289)
• ✨ Feat: Token Speed Display Feature @LazyYuuki (#5278)
• ✨enhancement: out of context troubleshooting @louis-menlo (#5275)
• ✨enhancement: model run improvement @louis-menlo (#5268)
• ✨enhancement: masking sensitive values in environment variables @urmauur (#5276)
• feat: improve ui search @samhvw8 (#5052)
• feat: hide docs link @samhvw8 (#5240)
• feat(local-api-server): add button to open API documentation and improve layout @samhvw8 (#5238)
• feat: render mcp content @samhvw8 (#5229)
• enhancement: mcp server activation response and error handling @louis-menlo (#5220)
• enhancement: tmp hide attachment icon @urmauur (#5160)
• enhancement: open folder log and change data folder dialog confirm @urmauur (#5159)
• enhancement: ux tool call permission dialog and state active @urmauur (#5157)
• enhancement: querry search hub include list of model @urmauur (#5147)
• enhancement: all platform layout @urmauur (#5119)
• enhancement: make the model selector popup responsive and wider for bigger screens @ethanova (#5025)
• enhancement: wrap long line codeblock @urmauur (#5069)
• enhancement: show assistant info on message @urmauur (#5064)
• enhance: setup screen redirect to hub @urmauur (#5061)
• enhancement: disabled config when server is running @urmauur (#5058)
• enhancement: update layout hub when left panel collapse @urmauur (#5059)
• enhancement: message toolbar using tooltip @urmauur (#5055)
• enhancement: built-in custom emoji and show metadata message @urmauur (#5085)
• feat: store last use model create thread @urmauur (#5201)
• feat: integrate fuzzy search into model dropdown @samhvw8 (#5197)
• feat: Jan API Server should have API Key setting @louis-menlo (#5193)
• feat: setting toggle vulkan @urmauur (#5126)
• feat: handle open Jan on HF GGUF repo @louis-menlo (#5173)
• feat: migrate legacy local storage data to new app @louis-menlo (#5156)
• feat: add refresh button list model remote provider @urmauur (#5136)
• feat: start and stop model @urmauur (#5133)
• feat: predefined params @urmauur (#5128)
• feat: filter downloaded model on hub screen @urmauur (#5113)
• feat: Support download resume @gau-nernst (#5111)
• feat: add quick access model setting via dropdown model @urmauur (#5104)
• feat: add couple section about us on general settings @urmauur (#5098)
• feat: product analytic @urmauur (#5099)
• feat: file upload thumbnail @urmauur (#5088)
• feat: improve download extension @gau-nernst (#5073)
• feat: Hardware info replacement for cortex @gau-nernst (#4925)
• feat: Implement Cortex server auto-restart and webview notification @samhvw8 (#5074)
• feat: add assistant emoji picker @urmauur (#5057)
• feat: initial app logs @urmauur (#5060)
• feat: add custom OpenAI provider @louis-menlo (#5033)
• feat: Jan supports multiple assistants @louis-menlo (#5024)
• feat: handle stop streaming message, scroll to bottom and model loads @louis-menlo (#5023)
• feat: add MCP server connection status @louis-menlo (#5018)
• feat: new frontend with model download function @louis-menlo (#5008)
• feat: tool use render UI @urmauur (#5004)
• feat: Download manager for llama.cpp extension @gau-nernst (#4933)
• feat: Add Tauri as an Alternative Build Option for Jan @louis-menlo (#4844)

🐛 Fixes

• fix(server): enhance CORS handling for local API network access @samhvw8 (#5236)
• Fix: fuzzy search model case sensitive @samhvw8 (#5228)
• fix: should not show vulkan options in cpu only mode @louis-menlo (#5210)
• Fix: mcp server file system missing param @david-menloai (#5204)
• Fix: headers validations for local api server @david-menloai (#5186)
• fix: relocate jan data folder @louis-menlo (#5179)
• fix: thread update after chat @samhvw8 (#5180)
• fix: stop api server on page unload @louis-menlo (#5356)
• fix: built in models capabilities @urmauur (#5354)
• fix: glitch toggle gpus @urmauur (#5353)
• 🐛fix: setting provider hide model capabilities @urmauur (#5355)
• 🐛 Fix: window path name with a space lead to cannot find correct path to Jan datafolder @LazyYuuki (#5351)
• fix(proxy): implement true HTTP streaming for chat completions API @samhvw8 (#5350)
• fix(server): add gzip decompression support for /models endpoint filtering @samhvw8 (#5349)
• 🐛fix: render streaming token speed based on thread ID & assistant metadata @urmauur (#5346)
• feat(server): filter /models endpoint to show only downloaded models @samhvw8 (#5343)
• 🐛fix: handle NaN value tokenspeed @urmauur (#5339)
• 🐛fix: update wording desc vulkan @urmauur (#5338)
• Fix: sorted thread not stable @samhvw8 (#5336)
• 🐛 fix: thread sorted not correctly after reload @samhvw8 (#5326)
• 🐛 fix(updater): terminate sidecar processes before update to avoid file access errors @Minh141120 (#5325)
• fix: glitch model download issue @louis-menlo (#5322)
• 🐛fix: allow script posthog @urmauur (#5316)
• 🐛fix: update default extension settings @louis-menlo (#5315)
• 🐛fix: llama.cpp default NGL setting does not offload all layers to GPU @louis-menlo (#5310)
• 🐛fix: safely convert values to numbers and handle NaN cases @urmauur (#5309)
• 🐛fix: immediately show download progress @urmauur (#5308)
• 🐛fix: glitch padding speed token @urmauur (#5307)
• 🐛fix: hide token speed when assistant params stream false @urmauur (#5302)
• 🐛fix: remove onboarding local model and hide the edit capabilities model @urmauur (#5301)
• 🐛fix: typo in build type check @louis-menlo (#5297)
• ✨enhancement: masking sensitive values in environment variables @urmauur (#5276)
• fix: glitch download from onboarding @urmauur (#5269)
• 🐛fix: showing release notes for beta and prod @urmauur (#5292)
• 🐛fix: optional additional information gpu @urmauur (#5291)
• 🐛fix: Jan-nano repo name changed @louis-menlo (#5274)
• 🐛fix: default onboarding model should use recommended quantizations @louis-menlo (#5273)
• fix: thinking block @urmauur (#5263)
• fix: default settings should leave empty @louis-menlo (#5257)
• fix: add cortex_killed_intentionally state to manage sidecar process termination @samhvw8 (#5255)
• fix: recheck tools approval state @urmauur (#5259)
• fix: replace close think tag from text segment @urmauur (#5258)
• fix: reduce app state left panel @urmauur (#5256)
• fix: cleanup console error and warning render FE @urmauur (#5253)
• fix: avoid persisting threads and messages on local storage @louis-menlo (#5249)
• fix: minor ui @urmauur (#5247)
• fix: tool call block width code @urmauur (#5246)
• fix: update style cortex dialog @urmauur (#5245)
• fix: update hgf jan nano @urmauur (#5242)
• fix: could not add custom models @louis-menlo (#5241)
• fix: tauri path env @louis-menlo (#5233)
• fix: MCP server should not spawn shell window @louis-menlo (#5223)
• fix: clean up cortex processes @louis-menlo (#5215)
• fix: tools call available dropdown @urmauur (#5222)
• fix: ui glitch dots truncated @urmauur (#5219)
• fix: conditional button refresh model @urmauur (#5221)
• fix: glitch UI thinking or duplicate content when multiple think after tools call @urmauur (#5217)
• fix: prettify json when pasting content @urmauur (#5216)
• fix: avoid spamming server restarts @louis-menlo (#5212)
• fix: path resolver in windows @louis-menlo (#5209)
• fix: load threads from api instead of states @louis-menlo (#5206)
• fix: force provider avatar to render in high quality @urmauur (#5203)
• fix: migrate provider settings @louis-menlo (#5205)
• fix: dropdown model position @urmauur (#5199)
• fix: truncate tool name @urmauur (#5198)
• fix: add missing default host ip @david-menloai (#5196)
• fix: remind me later updater @urmauur (#5191)
• fix: hub refresh state @urmauur (#5190)
• fix: handle tool availability states @louis-menlo (#5183)
• fix: avatar assistants render @urmauur (#5181)
• fix: migration should wait for extensions load @louis-menlo (#5177)
• fix: handle undefined search index in getFilteredThreads @samhvw8 (#5175)
• fix: model run failed on windows @louis-menlo (#5168)
• fix: app updater state @urmauur (#5171)
• fix: sticky action scroll to bottom when edit message @urmauur (#5169)
• fix: Fix hwinfo bugs @gau-nernst (#5164)
• fix: create new provider should not have previous settings @louis-menlo (#5166)
• fix: race condition get Assistant @urmauur (#5161)
• fix: remove capitalize class from provider name @urmauur (#5158)
• fix: extension settings are not retained in new sessions @louis-menlo (#5154)
• fix: avoid floating content when render table markdown @urmauur (#5148)
• fix: state model threadId @urmauur (#5144)
• fix: thread list state order after dragable @urmauur (#5141)
• fix: reverse setting local provider @urmauur (#5140)
• fix: content overflow when long text without space @urmauur (#5142)
• fix: tool call block spacing @urmauur (#5139)
• fix: support delete file for fs.rm() @gau-nernst (#5117)
• fix: chore UI issues @louis-menlo (#5116)
• fix: correct app data folder name @louis-menlo (#5115)
• fix: remove chat message on GUI @louis-menlo (#5114)
• fix: overflow content codeblock @urmauur (#5110)
• fix: prevent send message when empty message @urmauur (#5109)
• fix: enhance tool use and model provider not persisted issues @louis-menlo (#5094)
• fix: ci remove pr review trigger @Minh141120 (#5105)
• fix: fileStat returned object in JS @gau-nernst (#5102)
• fix: broken ui download count when leftpanel close @urmauur (#5089)
• fix: text component scaling @urmauur (#5090)
• fix: model import name issues @louis-menlo (#5093)
• fix: copy:lib on macos when directory already exists @gau-nernst (#5084)
• fix: provider activation status @louis-menlo (#5081)
• fix: bundle libraries on Linux and Windows @vansangpfiev (#5078)
• fix: mcp info when long string without space @urmauur (#5080)
• fix: hidden date when value 0 @urmauur (#5079)
• fix: close side windows should not kill background processes @louis-menlo (#5076)
• fix: allow dialog triggered by tooltip @urmauur (#5066)
• fix: tauri updater @vansangpfiev (#5051)
• fix: windows installer and allowed origins @vansangpfiev (#5037)
• fix: should not spawn many llama.cpp servers for the same model @louis-menlo (#4994)
• fix: solid background for linux and web @urmauur (#5163)
• fix: janhq url correction @david-menloai (#5107)
• fix: copy icon indicator when success @urmauur (#5091)
• fix: thread sorted and drag drop @samhvw8 (#5046)
• fix: type rename thread @samhvw8 (#5030)

🧰 Maintenance

• refactor: clean up repo @louis-menlo (#5165)
• chore: sync dev to release/v0.5.18 @louis-menlo (#5106)
• Update README.md @eckartal (#5072)
• chore: handle local models chat with MCP @louis-menlo (#5065)
• chore: add relocate jan data folder function to new FE @louis-menlo (#5043)
• chore: enable test on PR to release branch @vansangpfiev (#5028)
• chore: cleanup download management @urmauur (#5012)
• chore: Jan with new UI/UX @urmauur (#4964)
• chore: initial assistant feature @urmauur (#5013)
• Sync 0.5.18 to 0.6.0 @louis-menlo (#5320)
• Sync Release/v0.6.0 into dev @louis-menlo (#5293)
• Sync Release/v0.5.18 into Development branch @louis-menlo (#5213)
• chore: onboarding local model @urmauur (#5234)
• chore: correct binary name for stable version (#5303) @louis-menlo (#5311)
• chore: sync new hub data @louis-menlo (#5262)
• chore: enable shortcut zoom @urmauur (#5261)
• ci: add manual sign for appimage artifact @Minh141120 (#5254)
• chore: fix wrong classname color text @urmauur (#5252)
• chore: enhance hub guide @david-menloai (#5244)
• chore: experimental tag mcp @urmauur (#5239)
• chore: streaming tool output @urmauur (#5237)
• chore: fix model settings are not applied accordingly on change @louis-menlo (#5231)
• chore: add toggle loading sever MCP @urmauur (#5225)
• chore: build config info plist @Minh141120 (#5224)
• chore: force windows to default alpha color 1 @urmauur (#5218)
• chore: trim all mcp server configs before saving @louis-menlo (#5211)
• ci: update file size tauri build template macos and windows @Minh141120 (#5208)
• chore: thinking block auto collapse when streaming finish @urmauur (#5202)
• chore: update MCP servers list @louis-menlo (#5195)
• refactor: remove JS server package @louis-menlo (#5192)
• chore: hide some model capabilities @urmauur (#5189)
• refactor: clean up migrations @louis-menlo (#5187)
• ci: tauri build macos @Minh141120 (#5184)
• chore: enhance MCP permission settings @louis-menlo (#5178)
• chore: check app update via general setting @urmauur (#5167)
• refactor: deprecate legacy packages and clean up build scripts @louis-menlo (#5162)
• ci: add tauri build nightly manual @Minh141120 (#5153)
• ci: deprecate electron build @Minh141120 (#5149)
• chore: render markdown on user message @urmauur (#5146)
• chore: add timeout to tool use requests @louis-menlo (#5143)
• chore: hide reveal password @urmauur (#5138)
• chore: allow users to setting model offload @louis-menlo (#5134)
• chore: add skip button joyride @urmauur (#5125)
• chore: persist assistants settings @louis-menlo (#5127)
• chore: bump llama.cpp to b5509 @louis-menlo (#5120)
• chore: update make clean @gau-nernst (#5112)
• chore: add function to model settings @louis-menlo (#5108)
• chore: allow users to import hf repo @louis-menlo (#5103)
• chore: remove model provider from setup screen @urmauur (#5101)
• chore: translation @urmauur (#5097)
• chore: handle many issues with app settings and message actions @louis-menlo (#5086)
• chore: providers should default on @louis-menlo (#5083)
• chore: update font-src @vansangpfiev (#5077)
• chore: send chat completion with messages history @louis-menlo (#5070)
• chore: add model information in System Monitoring @louis-menlo (#5062)
• chore: wildcard for connect-src @vansangpfiev (#5048)
• chore: app updater UI @urmauur (#5054)
• chore: update Jan change logs v0.5.17 @david-menloai (#5056)
• chore: enable / disable proxy configrations @louis-menlo (#5050)
• chore: intial new window system monitor @urmauur (#5049)
• chore: add inline message chat input @urmauur (#5044)
• chore: show model name as filename @urmauur (#5045)
• chore: Jan's code is now under the Apache license @louis-menlo (#5042)
• chore: handle hardware settings @louis-menlo (#5041)
• chore: model import from llama.cpp provider @urmauur (#5038)
• chore: show location data folder and ui for let user change folder @urmauur (#5035)
• chore: fix auto focus chat input @urmauur (#5036)
• chore: token speed and edit message @louis-menlo (#5031)
• chore: model loading and truncate dropdown assiatant title @urmauur (#5032)
• chore: add setting hardware @urmauur (#5029)
• chore: remove electron nightly/manual @vansangpfiev (#5027)
• chore: reduce state update to prevent glitchy scroll to bottom @louis-menlo (#5026)
• chore: scroll to bottom when have new log message @urmauur (#5021)
• chore: regroup codeblock setting appearance @urmauur (#5022)
• chore: add function delete and params type assistant @urmauur (#5020)
• chore: stream app logs to log window @louis-menlo (#5019)
• chore: add some ts-ignore to make tauri build works @vansangpfiev (#5010)
• chore: cleanup whitespace left panel on thread list @urmauur (#5017)
• chore: improve switch component offset @urmauur (#5016)
• chore: allow users to enable/disable MCP servers @louis-menlo (#5015)
• chore: Pin rmcp commit @gau-nernst (#5014)
• chore: initial assistant feature @urmauur (#5013)
• chore: update model handlers on the new frontend @louis-menlo (#5011)
• chore: handle chat functions @louis-menlo (#5009)
• chore: update empty state thread list, search, and no model @urmauur (#5007)
• chore: initial setup screen for remote provider @urmauur (#5005)
• chore: increase hight of drag area window @urmauur (#4997)
• chore: delete Tauri pre-install in make clean @gau-nernst (#5000)
• refactor: remove hardcoded provider names @louis-menlo (#4995)

Contributor

@LazyYuuki, @Minh141120, @david-menloai, @eckartal, @ethanova, @gau-nernst, @louis-menlo, @ramonpzg, @samhvw8, @urmauur and @vansangpfiev @gunesacar @laurian19 @qnixsynapse