0.5.7-rc2-beta

What's Changed

• ci: disable turbo cache by @hiento09 in #3836

Full Changelog: v0.5.7-rc1-beta...v0.5.7-rc2-beta

0.5.7-rc1-beta

What's Changed

• test: add model parameter validation rules and persistence tests by @louis-jan in #3618
• Add support for 'o1-preview' and 'o1-mini' models by @RONNCC in #3659
• fix: phi3 medium prompt template by @louis-jan in #3670
• fix: #3667 - The recommended label should be hidden by @louis-jan in #3687
• fix: #3476 - Mismatch id between model json and path by @louis-jan in #3645
• Fix: #1142 setting groups toggle does not turn off it's nested settings by @louis-jan in #3681
• fix: small leftover issues with new starter screen by @urmauur in #3661
• feat: textarea auto resize by @urmauur in #3695
• chore: Copyfix by @0xSage in #3694
• fix: update the condition for generating the title by @urmauur in #3702
• fix: enhance various minor UI elements by @urmauur in #3706
• test: add web helpers, services, utils tests by @louis-jan in #3669
• fix: thread title for remote model from first prompt by @urmauur in #3712
• fix: remove title local API server page by @urmauur in #3710
• test: update test coverage UI component Joi by @urmauur in #3707
• test: add missing tests to Jan core modules by @louis-jan in #3716
• fix: #3693 broken thread.json should not break the entire threads by @louis-jan in #3709
• fix: toolbar overlap chat input by @urmauur in #3720
• chore: The monorepo jest configs should not cover extensions - they would be moved to mini repositories soon by @louis-jan in #3723
• fix: #3558 wrong model metadata import or download from HuggingFace by @louis-jan in #3725
• fix: #3513 - Anthropic extension does not forward the system prompt by @louis-jan in #3724
• fix: #3515 - The default assistant instructions are ignored by @louis-jan in #3721
• fix: #3698 - o1 preview models do not work with max_tokens by @louis-jan in #3728
• enhance: tabs component in model selection by @urmauur in #3730
• fix: #3673 - API responds with Request body is too large by @louis-jan in #3729
• fix: #3727 LLM model download fail can still be used by @louis-jan in #3731
• fix: #3703 - Deepseek-Coder-33B-Instruct is incompatible by @louis-jan in #3732
• fix: wrong model download location when there is a mismatch model_id by @louis-jan in #3733
• ci: auto trigger jan docs ci for new release by @hiento09 in #3734
• fix: #3491 - Unable to use tensorrt-llm by @louis-jan in #3741
• fix: correct model dropdown for local models by @louis-jan in #3736
• fix: model dropdown should show recommended models to download by @louis-jan in #3742
• fix: XML not render correctly by @urmauur in #3743
• chore: improve models and threads caching by @louis-jan in #3744
• release: Jan Release Cut v0.5.5 by @louis-jan in #3746
• fix: Optimize resource watching by @louis-jan in #3747
• chore: add llama3.2 and qwen2.5 models by @louis-jan in #3751
• fix: overlap text on model selector and loader by @urmauur in #3757
• fix: error handling for model imports should be handled gracefully by @louis-jan in #3763
• chore: move Jan Docs back into Jan Repo by @hiento09 in #3790
• fix: update base url edit page docs by @urmauur in #3795
• ci: Update jan-docs.yml by @hiento09 in #3799
• feat: save instruction for new threads as a stable feature by @urmauur in #3798
• fix: issue ellipsis text on model selector and model start loader by @urmauur in #3802
• test: added more test inside containers component by @urmauur in #3765
• fix: reset window state after disconnected from multi monitor by @urmauur in #3797
• fix: enable scroll modal component by @urmauur in #3816

New Contributors

• @RONNCC made their first contribution in #3659

Full Changelog: v0.5.4...v0.5.7-rc1-beta

0.5.6

Changes

• hotfix: graceful error handling model import @louis-jan (#3766)

Contributor

@louis-jan

0.5.5

Changes

• enhance: tabs component in model selection @urmauur (#3730)

🚀 Features

• feat: textarea auto resize @urmauur (#3695)

🐛 Fixes

• fix: thread title for remote model from first prompt @urmauur (#3712)
• fix: #1142 setting groups toggle does not turn off it's nested settings @louis-jan (#3681)
• fix: overlap text on model selector and loader @urmauur (#3756)
• fix: XML not render correctly @urmauur (#3743)
• fix: model dropdown should show recommended models to download @louis-jan (#3742)
• fix: correct model dropdown for local models @louis-jan (#3736)
• fix: #3491 - Unable to use tensorrt-llm @louis-jan (#3741)
• fix: wrong model download location when there is a mismatch model_id @louis-jan (#3733)
• fix: #3703 - Deepseek-Coder-33B-Instruct is incompatible @louis-jan (#3732)
• fix: #3727 LLM model download fail can still be used @louis-jan (#3731)
• fix: #3673 - API responds with Request body is too large @louis-jan (#3729)
• fix: #3698 - o1 preview models do not work with max_tokens @louis-jan (#3728)
• fix: #3515 - The default assistant instructions are ignored @louis-jan (#3721)
• fix: #3513 - Anthropic extension does not forward the system prompt @louis-jan (#3724)
• fix: #3558 wrong model metadata import or download from HuggingFace @louis-jan (#3725)
• fix: toolbar overlap chat input @urmauur (#3720)
• fix: #3693 broken thread.json should not break the entire threads @louis-jan (#3709)
• fix: remove title local API server page @urmauur (#3710)
• fix: enhance various minor UI elements @urmauur (#3706)
• fix: update the condition for generating the title @urmauur (#3702)
• fix: small leftover issues with new starter screen @urmauur (#3661)
• fix: #3476 - Mismatch id between model json and path @louis-jan (#3645)
• fix: #3667 - The recommended label should be hidden @louis-jan (#3687)

🧰 Maintenance

• chore: add llama3.2 and qwen2.5 models @louis-jan (#3750)
• chore: improve models and threads caching @louis-jan (#3744)
• ci: auto trigger jan docs ci for new release @hiento09 (#3734)
• chore: Copyfix @0xSage (#3694)

Contributor

@0xSage, @hiento09, @louis-jan and @urmauur

0.5.4

🚀 Features

• feat: enable search model dropdown by configured model @urmauur (#3466)
• feat: add setting options to dropdown menu @urmauur (#3554)

🐛 Fixes

• fix: correct prompt template for Phi3 Medium model @louis-jan (#3665)
• fix: change Phi-3 Mini Instruct 4K to cortexso repository @louis-jan (#3655)
• fix: prevent value empty string on slider right panel @urmauur (#3635)
• fix: #3549, #3552 - Inference on CPU is slower on Jan 0.5.3 @louis-jan (#3602)
• fix: missing type number for slider input right panel @urmauur (#3609)
• fix: disabled UI RAG and tools @urmauur (#3514)
• fix: using type text instead number for avoid region issue @urmauur (#3538)
• fix: electron window @urmauur (#3480)

🧰 Maintenance

• test: add tests to model setting's Slider Input @louis-jan (#3615)
• test: finished all core UI components test @urmauur (#3588)
• chore: update copy when retrieval turn off @urmauur (#3585)
• test: add core modules test cases @louis-jan (#3498)
• chore: setup jest for unit test hooks and component from joi @urmauur (#3540)
• chore: copy edit for model setting @urmauur (#3553)
• test: add test:coverage step @louis-jan (#3551)

Contributor

@0xSage, @louis-jan and @urmauur

0.5.3

Changes

• fix: gpu memory utilization @louis-jan (#3492)
• Release Cut 0.5.3 @louis-jan (#3473)
• Chore upgrade cortex 0.5.0 @hiento09 (#3462)
• feat: preserve model settings @louis-jan (#3427)
• Fix auto updater nightly @hiento09 (#3424)
• Chore use aws s3 for nightly channel @hiento09 (#3416)
• Chore/simple template report @Van-QA (#3400)
• Revert "Jan integrates Cortex" @louis-jan (#3371)
• Chore/simple bug template and correct a copy @Van-QA (#3344)
• Increase retry upload to R2 to 5 times @hiento09 (#3337)
• fix: remote engine should not allow reinit @louis-jan (#3284)
• chore: remove focus button @urmauur (#3272)
• bump-cortex-0.5.0-1 @Van-QA (#3218)
• Deprecate Docker and K8s @hiento09 (#3219)
• Fix typo Extention -> Extension @niesink (#3187)

🚀 Features

• fix: search input clearable component @urmauur (#3465)
• feat: add checklist regression test @Van-QA (#3425)
• feat: import model with gguf metadata extraction @louis-jan (#3421)
• feat: shortcut delete and clean thread @urmauur (#3423)
• feat: openrouter model settings @louis-jan (#3422)
• feat: add blank state component for hub and my model search @urmauur (#3413)
• feat: enable copy instruction @urmauur (#3401)
• feat: small ui improvement @urmauur (#3388)
• feat: allow user configure remote model from my model @urmauur (#3348)
• feat: move icon create new thread into top panel @urmauur (#3346)
• chore(UI): update experience model dropdown @urmauur (#3342)
• feat: modal waiting cortex @urmauur (#3306)
• feat: change data folder @louis-jan (#3309)
• feat: embed cortex into jan as a JS module @louis-jan (#3305)
• feat: add chunk count @namchuai (#3290)
• feat: add nvidia engine @namchuai (#3279)
• feat: enable copy over instructions @namchuai (#3266)
• feat: add more options for cortex popup @namchuai (#3236)
• feat: manage cloud models from threads screen @urmauur (#3223)
• feat: make scroll area type auto for make default visible scrollbar @urmauur (#3220)
• feat: new starter screen @urmauur (#3217)
• feat: integrating cortex @namchuai (#3001)

🐛 Fixes

• fix: responsive modal import gguf @urmauur (#3484)
• fix: add engine in payload for nitro chat completion api @marknguyen1302 (#3482)
• fix: wrong classname icon @urmauur (#3474)
• fix: modal responsive import model @urmauur (#3467)
• fix: resize left and panel when scroll area show @urmauur (#3470)
• fix: search input clearable component @urmauur (#3465)
• fix: import model by folder on win @marknguyen1302 (#3463)
• fix: correct time-weighted retrieval toggle @louis-jan (#3456)
• fix: add a minimum height for the main window otherwise it shrinks @louis-jan (#3457)
• fix: modal title should not truncated @urmauur (#3450)
• fix: responsive modal import model @urmauur (#3449)
• fix: enable extensions migration @louis-jan (#3442)
• fix: unable to import model due to missing metadata @louis-jan (#3439)
• fix: setting extension placeholder and input action apikey @urmauur (#3430)
• fix: preserve model settings should maintain all settings @louis-jan (#3435)
• fix: shortcut key on windows @urmauur (#3437)
• fix: date format for chat item @urmauur (#3436)
• fix: setting chatbox input @urmauur (#3429)
• fix: missing font katex @urmauur (#3431)
• fix: decreased scrollbar width @urmauur (#3432)
• fix: add toaster warning when blocked create new thread @urmauur (#3433)
• fix: save state toggle input box setting @urmauur (#3402)
• fix: avoid show remote model on system monitor @urmauur (#3412)
• fix: disable remote model when user did not setup apiKey @urmauur (#3415)
• fix: app emits ugly errors when retrieving settings @louis-jan (#3393)
• fix: Jan should have access to it resource folders @louis-jan (#3391)
• fix: LaTex formula render issue @louis-jan (#3353)
• fix: failed to relaunch app to update @louis-jan (#3351)
• fix: empty model page not shown when delete all threads and models @namchuai (#3343)
• feat: allow user configure remote model from my model @urmauur (#3348)
• fix: broken jan build - add log trace (jan.log) @louis-jan (#3336)
• fix: update legacy path @namchuai (#3328)
• fix: broken app due to incorrect api path @louis-jan (#3316)
• fix: refresh should not create new thread @namchuai (#3314)
• fix: avoid lose title threads @urmauur (#3307)
• fix: migration item in setting detail omit buttons @namchuai (#3298)
• fix: merge gpu arch and os tensorrt models @namchuai (#3299)
• fix: engine logo on model dropdown @urmauur (#3291)
• fix: icon setting can close and open right panel @urmauur (#3295)
• fix: error while importing local model is not shown @namchuai (#3294)
• fix: add lower case quantization support @namchuai (#3293)
• fix: onnx can't be selected in download model modal @namchuai (#3283)
• fix: system bar always show RAM usage 0% @louis-jan (#3287)
• fix: migration wrong directory @namchuai (#3278)
• fix: clearer app loading prompt @louis-jan (#3275)
• fix: allow user to reinit engine from settings page @louis-jan (#3277)
• fix: should not let second instance terminate cortex @louis-jan (#3274)
• fix: set cortex data folder path when starting jan @namchuai (#3252)
• fix: remote model added manually does not shown in model drop down @namchuai (#3261)
• fix: normalize messages for some providers @namchuai (#3234)
• fix: broken status parse due to empty category @louis-jan (#3233)
• fix: update new api from cortex to support 0.5.0 @namchuai (#3221)
• fix: jan app copy nits @0xSage (#3193)

🧰 Maintenance

• chore: Update model.json with correct filename @Van-QA (#3479)
• chore: add llama 3.1 8B gguf model @louis-jan (#3468)
• chore: move remote models to bottom of the models hub @louis-jan (#3469)
• chore: add husky pre-commit @louis-jan (#3458)
• chore: change jan default data folder path to app's userData @louis-jan (#3443)
• chore: add back OpenRouter request endpoint setting @louis-jan (#3441)
• chore: fix linter issues @louis-jan (#3434)
• chore: disable internal REST endpoints due to a vulnerability issue @louis-jan (#3428)
• chore: small ui improvement @urmauur (#3414)
• chore: fix CI - remove headless ui @louis-jan (#3420)
• chore: remove model selection in assistant tab model @urmauur (#3411)
• chore: clean up console modal component @urmauur (#3395)
• feat: add model assistant tab @urmauur (#3389)
• chore: improve readability with set max width container message @urmauur (#3382)
• feat: small ui improvement @urmauur (#3388)
• chore: move create new thread to top panel @urmauur (#3381)
• chore: update model dropdown and my model with collapsible chevron @urmauur (#3386)
• chore: cherry pick latest fixes @louis-jan (#3383)
• chore: add engine logo from local instead of metadata logo @urmauur (#3363)
• chore: minor ui improvement @urmauur (#3352)
• chore: add back GPU information to system monitoring bar @louis-jan (#3350)
• chore: button outline for configure cloud model @urmauur (#3347)
• chore(UI): update experience model dropdown @urmauur (#3342)
• chore(ui): fix alignment loader starter screen @urmauur (#3338)
• chore: disable quick ask @louis-jan (#3334)
• chore: add cortex version @namchuai (#3318)
• chore: update cortex new version @namchuai (#3300)
• chore: update minor UI @urmauur (#3281)
• chore: update z-index tooltip @urmauur (#3280)
• chore: toast message on model import fail with reason @louis-jan (#3276)
• chore: update hub search result @urmauur (#3273)
• chore: temporary hidden button import model @urmauur (#3270)
• chore: check the legacy incompatible message type @louis-jan (#3248)
• chore: minor copy for grammar @gabrielle-ong (#3235)
• chore: update conditional starter screen after cortex load @urmauur (#3227)
• chore: set container max width for chat message and new hub screen @urmauur (#3213)

Contributor

@0xSage, @Van-QA, @gabrielle-ong, @hiento09, @louis-jan, @marknguyen1302, @namchuai, @niesink and @urmauur

0.5.2

Changes (#2999)

• chore: Bump-cortex-0.4.17 (#3111)
• chore: cortex version update (#3098)
• chore: update download.ts (#3088)
• chore: upgrade marked-katex-extension (#3049)

🚀 Features

• feat: add claude 3.5 sonnet (#3129)
• feat: add options config spell check for chat input (#3131)
• feat: add shortcut app logs on system monitor (#3150)
• feat: enable right click to show setting on thread items (#3108)

• chore: release 0.5.2 to main @Van-QA (#3158)

🐛 Fixes

• fix: add tooltip messages toolbar (#3138)
• fix: handle error message when apikey is not setup (#3149)
• fix: handle long thread title without space (#3107)
• fix: handle long word without space to avoid right panel disappears (#3048)
• fix: handle words without space (#3101)
• fix: missing keyboard sand shortcuts screen (#3156)
• fix: model dropdown search by configured model (#3047)
• fix: model dropdown search result always show downloaded models first (#3147)
• fix: move Jan folder (#3160)
• fix: set specific version for terminate (#3126)
• fix: title thread not updated on input edit title (#3148)
• fix: toggle button for expand log section on modal troubleshoot (#3130)
• fix: validate path in fs (#3152)

🧰 Maintenance

• Update cortex cpp nightly to version 0.4.18 (#3072)
• Update cortex cpp nightly to version 0.4.18 (#3114)
• add time weighted retrieval (#2908)
• bump version (#3082) (#3159)
• Chore: Add stop token for Gemma 2b (#3125)
• fixed grammar nits (#3132)

Contributor

@0saurabh0, @Van-QA, @eltociear, @github-actions[bot], @hahuyhoang411, @jan-service-account, @marknguyen1302, @namchuai, @louis-jan, @thunhuanh and @urmauur

0.5.1

Changes

• Chore: Model bump version @hahuyhoang411 (#3082)
• Chore: ngl adjust for correct usage @hahuyhoang411 (#3081)
• Fix anthropic streaming error @namchuai (#3079)
• Revert "Update cortex cpp nightly to version 0.4.12-16.06.24" @hiento09 (#3052)
• Change reviewer for nightly PR update cortex cpp @hiento09 (#3046)
• Update cortex cpp nightly to version 0.4.12-16.06.24 @jan-service-account (#3045)
• Add nightly update cortex cpp to Jan app @hiento09 (#3044)
• Chore: model hub v0.5.1 update @hahuyhoang411 (#3036)
• Added NVIDIA API to new jan after jan rework @Realmbird (#2934)
• Sync main to dev 0.5.0 @Van-QA (#2984)

🚀 Features

• fix: reduce font-weight for message @urmauur (#3018)
• feat: enable hotkey collapse right panel @urmauur (#3019)
• feat: update icon alert warning and copies tooltip model dropdown @urmauur (#3021)

🐛 Fixes

• fix: disable flash_attn if vulkan is enabled @namchuai (#3080)
• fix: position dropdown on my model @urmauur (#3058)
• fix: enable window control button on linux @urmauur (#3028)
• fix: reduce font-weight for message @urmauur (#3018)
• fix: bring back maximize button @urmauur (#3024)
• fix: duplicate role inside messages cause some model to refuse to answer @namchuai (#3006)
• fix: specific version for turbo @namchuai (#2989)
• fix: scan the models folder recursive to find model metadata file @namchuai (#2982)

🧰 Maintenance

• chore: release 0.5.1 to main @Van-QA (#3051)
• chore: update template check box @Van-QA (#3039)
• chore: Remove OpenAI test coverage @hiro-v (#3033)
• chore: Refactor/issue template feature request @Van-QA (#3037)
• chore: Refactor/issue template and extension version bump @Van-QA (#3035)
• chore: update some copy @namchuai (#3034)
• chore: Bump-cortex-0.4.13 @Van-QA (#3027)
• fix: loading jan browser @phoval (#3011)

Contributor

@Realmbird, @Van-QA, @github-actions[bot], @hahuyhoang411, @hiento09, @hiro-v, @jan-service-account, @namchuai, @phoval and @urmauur

0.5.0

Changes (#2904)

• Release/0.5.0 to main @Van-QA (#2981)
• Chore: Model Hub update @hahuyhoang411 (#2966)

🚀 Features

• feat: add input actions for setting item @namchuai (#2978)
• feat: update setting tooltip explainer @urmauur (#2972)
• feat: allow user to register their access token @namchuai (#2974)
• fix: adjustment minor UI from qa feedback @urmauur (#2963)
• feat: Dekstop Revamp @urmauur (#2877)

🐛 Fixes

• fix: quickask missing color schema @urmauur (#2979)
• fix: missing markdown list and blockquotes @urmauur (#2980)
• fix: glitch modal troubleshoot @urmauur (#2976)
• fix: theme migration overwrite existing theme @louis-jan (#2975)
• fix: clickable remote model @urmauur (#2970)
• fix: max value of max_tokens should follow context_length's value @namchuai (#2969)
• fix: wrong direction icon collapse right panel @urmauur (#2967)
• fix: adjustment minor UI from qa feedback @urmauur (#2963)
• fix: max_tokens and ctx_len max value will follow model.json @namchuai (#2961)

🧰 Maintenance

• chore: upgrade version model extension for hf auth token @namchuai (#2983)
• chore: Bump-cortex-0.4.11 @Van-QA (#2962)

Contributor

@Van-QA, @hahuyhoang411, @louis-jan, @namchuai and @urmauur

0.4.14

Changes

• Release/0.4.14 to main @Van-QA (#2950)
• Update package.json to use [email protected] @Van-QA (#2949)
• Chore: aya update @hahuyhoang411 (#2941)
• Code sign retry 3 times @hiento09 (#2943)
• Bump cortex to 0.4.9 @Van-QA (#2940)
• Update README.md @eckartal (#2927)
• Bump cortex to 0.4.8 @Van-QA (#2938)
• Chore: phi3 long-context update @hahuyhoang411 (#2936)
• Feat/oai endpoints mapper using JSON @Van-QA (#2929)

🐛 Fixes

• fix: unchange title and last message when clean or delete message @urmauur (#2937)
• fix: fix crash when model name not ready @namchuai (#2931)
• Fix: Phi-3 doesn't display @hahuyhoang411 (#2928)

🧰 Maintenance

• docs: Update README.md @Marco-9456 (#2939)
• Replace deprecated steps github action @hiento09 (#2935)

Contributor

@Marco-9456, @Van-QA, @eckartal, @hahuyhoang411, @hiento09, @namchuai and @urmauur