0.5.15

🚀 Features

• feat: Allow HTTP proxy authentication inputs @bxdoan (#4479)
• enhancement: remote engine model @urmauur (#4560)
• enhancement: error ui in threads @urmauur (#4578)
• fix: think content render as markdown @urmauur (#4564)
• feat: better hardware setting @urmauur (#4471)
• enhancement: update UI uploading files and images @urmauur (#4562)
• feat: pull models list when setting up a new remote engine @louis-menlo (#4547)
• feat: Standardize inline error messages @urmauur (#4530)
• enhancement: Inconsistent text color, hidden api-key field, chatcompletion field under advance setting @urmauur (#4532)
• enhancement: update responsive setting and my model page @urmauur (#4533)
• feat: allow users to add remote models @louis-menlo (#4534)
• enhancement: Update app layout with an expanded system monitor @urmauur (#4528)
• enhancement: added space scroll remote engine setting @urmauur (#4523)
• enhancement: create new thread with last chosen model @urmauur (#4524)
• feat: remove umami @urmauur (#4520)
• enhancement: minor ui refinements @urmauur (#4521)

🐛 Fixes

• fix: typo on apple @ynshung (#4525)
• fix: reload app when gpu change @urmauur (#4678)
• fix: app stucks at generating response state after stopped the previous session @louis-menlo (#4662)
• chore: disable turn off local engine llamacpp @urmauur (#4660)
• fix: link open new window inside chat message @urmauur (#4651)
• fix: gpu hardware state @urmauur (#4650)
• fix: error setting remote engine @urmauur (#4648)
• fix: model dropdown icon miss align @urmauur (#4645)
• fix: rename google gemini into gemini @urmauur (#4644)
• fix: multiple gpu @urmauur (#4641)
• fix: filter out variant engine @urmauur (#4639)
• fix: beta noti discord @hiento09 (#4640)
• fix: messages re-render on different threads @louis-menlo (#4638)
• fix: render slow on device label @urmauur (#4627)
• fix: app shows update button even there is no new release @louis-menlo (#4628)
• fix: remove toggle enable GPU from setting @urmauur (#4624)
• fix: filter out fake gpu @urmauur (#4617)
• fix: inconsistent models list on first launch @louis-menlo (#4618)
• fix: remove from list dropdown when deleted remote model @urmauur (#4616)
• fix: server log actions position @urmauur (#4608)
• fix: remote engine model auto select @urmauur (#4609)
• fix: load model error component @urmauur (#4604)
• fix: download progress import from hgf @urmauur (#4567)
• fix: think content render as markdown @urmauur (#4564)
• fix: app does not reload model when prompt template is updated @louis-menlo (#4537)
• fix: broken range slider ctx length @urmauur (#4535)
• fix: typo meta description @urmauur (#4507)

🧰 Maintenance

• chore: privacy policy final @eckartal (#4472)
• chore: fix engine template for remote providers @louis-menlo (#4680)
• chore: bump cortex release 1.0.10 @louis-menlo (#4679)
• chore: Sync Server Playground Schema with Latest Cortex.cpp API Specs @louis-menlo (#4672)
• chore: update ui width text field slider, and conditional button copy based value length @urmauur (#4661)
• chore: disable turn off local engine llamacpp @urmauur (#4660)
• chore: bump cortex version and fix misc UI issues @louis-menlo (#4654)
• chore: remote engine error handling @louis-menlo (#4646)
• chore: beta version not publish to github @hiento09 (#4634)
• chore: bump cortex.cpp to the rc5 to address race condition issue @louis-menlo (#4630)
• chore: fix engine variant for mac targets @louis-menlo (#4626)
• chore: add back env paths setting so cortex engine can load libraries properly @louis-menlo (#4625)
• chore: change cortex.cpp log level to info for easier log observation @louis-menlo (#4623)
• chore: bump cortex to the latest version 1.0.10-rc1 to address hardware UI issues @louis-menlo (#4622)
• fix: error handling - pull remote models list raise an error on fail @louis-menlo (#4615)
• chore: add back app logging function @louis-menlo (#4614)
• chore: update remote provider response transform template @louis-menlo (#4602)
• chore: bump cortex with conversation patch update @louis-menlo (#4598)
• chore: add google gemini and deepseek as new supported remote engine providers @louis-menlo (#4596)
• chore: Updated copies in-app @imtuyethan (#4527)
• chore:add deepseek r1 distill models to hub @louis-menlo (#4586)
• chore: add default inference parameters for custom remote models @louis-menlo (#4584)
• chore: calculate model label @urmauur (#4585)
• chore: engine model adding ux refinement @louis-menlo (#4580)
• chore: delete custom engine @urmauur (#4581)
• chore: reasoning block @urmauur (#4551)
• chore: align item menu action error message @urmauur (#4548)
• chore: ensure app Id is maintained @louis-menlo (#4541)
• chore: Update outdated integrations docs @imtuyethan (#4516)
• chore: don't auto-sign assignees from external sources due to security concerns @Minh141120 (#4506)

Contributor

@Minh141120, @bxdoan, @eckartal, @hiento09, @imtuyethan, @louis-menlo, @urmauur and @ynshung