Hub 0.1.27Latest
- Workshop builds reach your model again. A build could stop at once with "Failed to fetch" and leave the empty template.
- Cloud agents that use a model on this computer or your endpoint no longer stop with "Failed to fetch".
What's
new.
Every update to every Jennefer app, newest first, in plain words.
- 33Releases
- 9Apps
- Oct 22Last update
LGTM is earned, not given.
Patch, reviewer · tap me
Showing 33 releases from every app
October 2026
Hub 0.1.26
- Download bars on app cards fill from the left to the real percentage again, instead of showing a small block in the middle.
- Bars on light cards and in onboarding have a visible track, so you can see how far a download or model pull has come.
- Update, model and storage bars grow smoothly and start empty at 0%.
Hub 0.1.25
- Turn on "Use from your phone" for an endpoint model in Models > Your endpoint. Your phone chats and Cloud builds can then use it while this computer is off (Pro plan).
- Chats from your phone can use any model on this computer or your endpoint, with an effort level, and the phone shows when it waits for the GPU or loads a model.
- Models > Each app lists Mobile, so you can give your phone its own default model.
- Models marks the right Hub default and shows Loaded for a model that a session is using.
Workshop 0.1.16Latest
- Apps read the text of PDFs reliably, including long books with special fonts, through Jennefer instead of their own reader.
- Ask Jennefer to switch your app to the new PDF reading if it says a PDF has no text.
Workshop 0.1.15
- Pick which AI model an app uses, right in its top bar: for your own apps and for apps from Explore.
- Apps can keep files such as PDFs and images (up to 100 MB each), and show the images they save.
- Your own apps can save as much as your disk allows; Workshop only stops when less than 1 GB is free.
- Apps from Explore can use up to 500 MB, and Permissions shows how much they saved, with Clear data.
Workshop 0.1.14
- Apps can save much more: up to 2 MB per item and 50 MB per app, so apps that keep document or PDF text no longer hit a limit.
- When an app works with long text, Jennefer now splits it into parts instead of cutting it short.
- Saving is faster in apps with a lot of data. Data your apps already saved moves over by itself.
Workshop 0.1.13
- Explore: open apps other people made in Workshop. They run on this computer with your own AI model, and you choose what each app may do.
- Publish your apps to the gallery with Workshop Pro, All Apps Pro or a Team plan, and ship updates with notes.
- Apps from the gallery update by themselves, and ask first when an update needs more permissions.
- Report a gallery app, or remove it from this computer, from its top bar.
Workshop 0.1.12
- Add images and files to your request: use the paperclip, paste, or drag them into the box.
- Show a screenshot or sketch as a reference, and the builder uses it as a guide.
- Logos, photos and data files (CSV, JSON, text) go into your app, ready to use.
- Workshop tells you when the picked model cannot see images.
Workshop 0.1.11
- Drag and drop works inside the apps you build.
- Open the running app in your browser from the preview toolbar.
- Builds no longer stop with raw tool text in the chat; the change is made instead.
Mobile 0.2.0Latest
- Chat with any model on your computer or your endpoint, choose an effort level, and keep many chats with rename and delete.
- Attach photos, PDFs and text files. Code blocks have a Copy button, and every answer shows its tokens and cost.
- See when your computer is busy or loading a model instead of a blank wait.
- Keep an endpoint in your account (Pro) to chat with it while your computer is off.
- Build by chat: describe an app and an agent builds it in your Cloud sandbox, then open the preview on your phone.
Hub 0.1.24
- Choose where Jennefer keeps its files: Settings > Storage moves your models, Studio packs and the local runtime to another drive, and new apps install there too.
- Models shows where your models are saved and how much space is free.
- Model downloads show one steady progress bar instead of starting over for every part.
- App installs keep their progress when you switch screens, and the update card shows the right percentage.
- Your endpoint answers requests that carry only an instruction, so apps built in Workshop can use it.
Hub 0.1.23
- Your endpoint works with Claude on the Anthropic API and on Amazon Bedrock, with prompt caching. Save several endpoint models and choose an effort level for each.
- Give each app its own model in Models > Each app, and see under Open now which window or terminal uses which model.
- Usage has three tabs: this computer, your endpoint (estimated cost at list prices, with cache reads and writes) and Jennefer Cloud.
- GPU shows the model each job uses and when a model is loading. Apps that use different models on this computer take turns without switching models on every step.
IDE 0.1.14Latest
- Comes with the latest Jennefer Engine: your models can live on another drive (Jennefer Hub > Settings > Storage).
- Model downloads show steady progress.
- Agents keep going after they look something up instead of stopping early, and a missing program is reported instead of retried.
IDE 0.1.13
- Each window can use its own model and effort: open the model chip, pick one, or make it the IDE default.
- Endpoint models are shown in blue, with this window's estimated cost and cache use.
- A line above the message box tells you when your request waits for the GPU or the model is loading.
- Agents now see the error output of commands that succeed, and Claude models call their tools more reliably.
Studio 0.1.6Latest
- Studio packs can live on another drive: move them in Jennefer Hub > Settings > Storage.
- While Jennefer moves its files, Studio waits and tells you why.
Cloud 0.1.9Latest
- Comes with the latest Jennefer Engine: your models can live on another drive (Jennefer Hub > Settings > Storage).
- Model downloads show steady progress.
- With your endpoint the team uses the model's full context window and no longer re-reads the same files.
Cloud 0.1.8
- Pick the model and effort for this window next to the message box, or make it the Cloud default.
- A line tells you when the team's request waits for the GPU on your computer or the model is loading.
- Agents now see the error output of commands that succeed, and Claude models call their tools more reliably.
Workshop 0.1.10
- Comes with the latest Jennefer Engine: your models can live on another drive (Jennefer Hub > Settings > Storage).
- Model downloads show steady progress.
- Apps you build can use your endpoint model.
- With your endpoint, builds use the model's full context window and no longer re-read the same files.
Workshop 0.1.9
- Pick the model and effort for this window next to the build box, or make it the Workshop default.
- Each build shows its tokens and, on your endpoint, its estimated cost.
- While a build waits for the GPU or the model is loading, Workshop says so.
CLI 0.2.9Latest
- /model picks the model for this conversation (/model save makes it the CLI default), and /effort sets how much the model thinks.
- /cost shows tokens, cache use and the estimated cost on your endpoint. Endpoint models have their own color.
- The working line tells you when your request waits for the GPU or the model is loading.
- Cleaner output: no placeholder thoughts, and team commands show which agent runs them.
CLI+ 0.1.11Latest
- Comes with the latest Jennefer Engine: your models can live on another drive (Jennefer Hub > Settings > Storage).
- Model downloads show steady progress.
CLI+ 0.1.10
- /model and /effort pick the model and effort for this conversation; /model save makes it the CLI+ default.
- The HUD shows your endpoint model, its effort and the conversation's estimated cost. /cost has the details.
- A line above the input tells you when your request waits for the GPU or the model is loading.
Hangar 0.1.2Latest
- Comes with the latest Jennefer Engine: your models can live on another drive (Jennefer Hub > Settings > Storage).
- Model downloads show steady progress.
- Bot agent steps keep going after they look something up.
Hangar 0.1.1
- Each agent step can use its own model and effort: set it in the step's Model field. Steps without one use the Hangar default from the Hub.
- Choose Jen's model for this window under her message box.
- Running flights show up in Hub > Models > Open now with the bot's name.
Hub 0.1.22
- Meet Jennefer Hangar: bots that watch pages, prices, folders and Telegram chats for you. Install it from Home when it opens for your account.
- The tray menu can ground all your Hangar bots at once, and shows how many are launched.
- Use from your phone now also lets you answer and ground your Hangar bots from Jennefer Mobile.
Hub 0.1.21
- The model list now updates on its own when Jennefer adds or removes a model. You no longer wait for a new Hub version.
- Models shows which catalog version you have, with Check for updates.
- If the model you use is removed from the catalog, Models tells you and asks you to pick another one.
Hub 0.1.20
- Unload now checks that the model really left the GPU. If it stays in memory, the Hub tells you instead of saying it was unloaded.
Hub 0.1.19
- New Usage screen: see how many tokens each app used with the model on this computer, by day, week or month.
- Pick a cloud model such as Claude Sonnet 5.5 to see what the same work would have cost there.
- Only counts are kept, never what you wrote or what the model answered.
CLI+ 0.1.9
- Jennefer Hub > Usage now shows the tokens CLI Plus uses as its own line, next to the other apps.
- Pick a cloud model there to see what the same work would have cost; only counts are kept, never what you wrote.
Hangar 0.1.0
- Tell Jen what to keep an eye on, and Jen builds the bot: a plan you can read, test and launch.
- Bots fly on a schedule, when a Telegram message comes in, when a file lands in a folder, or when another bot calls them.
- Bots can read pages and APIs, answer from your documents, make images and ask you before anything risky.
- Everything runs on your computer with the model you picked in Jennefer Hub. Hangar never places trades.
CLI 0.2.8
- Jennefer can use your MCP servers. Type /mcp to see the servers you set up in Jennefer Hub > Settings > MCP.
- /mcp <name> turns a server on or off for the folder you are in. /mcp hub opens the Hub to add one.
- Jennefer asks you before every MCP tool call. Plan mode never uses them.
- After an update the CLI shows what is new, like this. /changelog shows it again.
CLI+ 0.1.8
- The crew can use your MCP servers. Type /mcp to see the servers you set up in Jennefer Hub > Settings > MCP.
- /mcp <name> turns a server on or off for this project. The crew asks you before every MCP tool call.
- After an update, CLI Plus shows what is new, like this.
Hub 0.1.18
- After each update the Hub shows what is new, like this.
- Update messages list what changed before you install.
- Models: if your GPU is nearly full, you can pick up to two bigger context windows after a warning. If the model does not load, the Hub switches back by itself.
- Models: when your GPU leaves little room for agents, the Hub suggests a smaller model that fits.
- You can choose to share anonymous usage statistics in Settings > Privacy. It stays off until you turn it on.