Bugs

Urgent

All bugs are scoped in major architecture revamp, but the following bugs might be considered as directly affecting user retention rate, it would be great if we could patch some quick fixes instead of waiting for big epics to be delivered:

• bug: Unsupported Parameter Errors for model o1-mini & o1-preview  #3771
• bug: Unsupported parameter: 'stop' is not supported with this model on o1 preview  #3745

Important

Other separate issues:

• Some issues in QA: Jan 0.5.7 Release Sign-off #3818 that Eng is not committed to delivering in 0.5.7, i will update later.
• bug: mac everytime i open the jan.app it just keeps jumping for few minutes and then opens #3657
• bug: Nemo model - Jan gets stuck on "Stopping model" after model settings change #3774
• bug: Some UI/UX issues with model import #3811
• bug: Jan is unable to quit when Quick Ask is activated #3679

Enhancement

Enhancement to the current feature set that addresses user's needs & improves user experience.

P0

The following enhancement can solve several bug reports:

• feat: Prompt user to download model, if no models are available #3537
• feat: Pick new model in thread if original model is unavailable #3385
• idea: Improve Time-weighted Retrieval? Remove it? #3619 (a community-built feature; no team member knows what it does).

P1

• epic: Better app updater (with changelog) #1471
• epic: Jan Context Length issues #2320
• feat: Jan can update Models #3251
• epic: add cache_prompt to Engines Settings #3320 (considering Anthropic reaches >20M users now, this could increase Jan's retention by allowing Claude users to save more money using Anthropic models in Jan).
• idea: Better Hardware Settings & Error Handling #3688
• feat: Show model ID in Local API Server #3752
• feat: User can use Stop Words presets and add custom #3536
• epic: Experimental features should have experimental tag #3096

P2

• feat: Add pause functionality to model downloader #3519
• feat: Support response_format and structured JSON responses. #3785
• idea: integrity check on a downloaded model #3775
• idea: [UX enhancement] Handle port conflicts gracefully in Jan API Server  #3814
• idea: [UX enhancement] Improve Context Length and Max Tokens Settings UX #3738
• idea: [UI enhancement] System monitor should not cover input box #3812
• feat: Users can open model, thread files directly from threads Right Panel #3758
• feat: right-click menu options for selected text (in threads conversations) #1101

Features

P0

P0 features are what @dan-homebrew and @0xSage already laid out:

• planning: Jan's path to cortex.cpp? #3690
• planning: Provider Abstraction #3786
• planning: Jan Architecture supports both Mobile and Desktop #3735
• planning: Users Want Advanced RAG #3550 (To resolve tons of reported bugs & user requests around RAG)
• feat: Jan supports new Cortex's Model Folder and model.yaml architecture #3633
• epic: Jan's Revamped Model Hub #3461
• planning: Jan's Extension Framework #3773
• epic: Jan has Web Search #3059
• planning: UI Extensions Framework #2718
• Multi-modal Support (vision, voice) epic: Jan supports Local Voice Mode w/ Ichijo #3488 (handled at Cortex level; Jan needs UI design).
• epic: Jan's Automated Testing Strategy #3632 (Better QA process, please!)

P1

Apart from P0 features, here are some features that tons of users are begging to have in Jan:

• feat: Opt-in Analytics Analytics Infra #3503 (Currently, we still use Umami to track DAU, but we didn't publicly explain it. Users come to us for privacy solutions & they are asking why their data is sent to Umami).
• epic: Better Threads Management #3133
• epic: Jan & Cortex support for AMD GPUs  #3529
• idea: Search within Jan app #3648
• epic: Jan has Model Presets #2845 (handled at Cortex level; Jan needs UI design).
• feat: Jan allows bulk import and subfolder recognition for model management #2843
• feat: users can edit AI responses inline in threads UI #2842
• feat: Ready-to-use document library #2761
• feat: Save system prompt to assistant(s) #2255

Some other P1 issues are handled at the Cortex level.