Changelog
What’s new
New models, smarter Auto and new features, newest first.
Math in your exports, upload progress, and a smarter Apier
- PDF, Word and PowerPoint files you ask for now show formulas as real math, with neater tables, code boxes, lists and page numbers.
- Uploads show a progress bar, then “reading…” while Apier reads the file.
- Ask any model which model it is, how many credits you have left, what a reply costs or how Auto and Team mode work: it now knows Apier and gives exact numbers.
- Long lecture decks reach the model whole, and if part of a file is cut, the reply says so plainly.
- Flashcards and quizzes from a folder list every file they used.
- Files are checked before they're opened, so a damaged or disguised file gets a clear message.
Practice quizzes from your folders
- Practice quiz and Flashcards on any Workspace folder: Apier builds them from that folder's notes, files and chats, in a new chat filed there.
- A small ring by the send button shows how much of this month's credits is left; tap it for a quick summary. You'll also get a heads-up when 80% is used.
- Long answers that reach the length limit now carry on by themselves instead of asking you to say continue.
More credits, and longer answers
- More credits on paid plans: Pro now has 15,000 credits a month (was 10,000) and Student 6,000 (was 5,000). If you're on a paid plan, this month's credits have been topped up too.
- Long chats use fewer credits: the earlier part of a conversation now costs about a tenth as much each time you send a message.
- Longer answers: replies have much more room before they're cut off, and essays, reports and stories get the most.
Workspace: folders for your chats
- Workspace in the sidebar: your folders and chats in one list, like a Drive, in a grid or a list.
- Make folders, and folders inside folders, with a name and a colour; pin the ones you use most to the sidebar.
- A chat can be in as many folders as you like. Drag chats onto a folder (from the sidebar too), or choose Add to folders from a chat's menu.
- One + button adds a folder, a new chat, existing chats or files, wherever you are.
- Search a folder's names, chats and everything said in them; filter by folders, chats, pinned or chats with files, and sort by last used or name.
- Inside a folder, every file and image from its chats is shown at the bottom, with a preview.
Rename your chats
- Rename any chat: double-click it or use the pencil on a computer; on a phone, press and hold it (or tap ⋯ on the open chat) and choose Rename.
Cleaner plans, memory and chat list
- Replies now read as plain text on the page, without a bubble around them. Your own messages keep theirs.
- A new Plans page: switch between monthly and yearly, see what each plan includes, and the upgrade buttons line up.
- A new Memory page: search, edit and forget memories more easily, on phones too.
- Deleting all memory now asks in a small confirmation box, instead of making you tap twice.
- Every panel closes the same way, with an ✕ in the top corner.
- On phones, the chat list no longer shows a delete ✕ on every chat. Press and hold a chat, or tap ⋯ on the open one, to delete it.
Redo, more credits, and a backup when a model fails
- Drop files anywhere on the screen to attach them, and pick any file type: Jupyter notebooks, Excel and PowerPoint files, and many more code formats are now read.
- Deleting a chat remembers your choice when you tick “Remember my choice”, and skips the question next time (change it in Settings → Memory).
- Stop now stops right away, even before the reply has started, instead of showing “Still writing this reply”.
- The live status shows how many tokens the reply has used so far, keeps a steady size, and lists each step the moment it starts.
- See the work: tap the status while a reply is being written to open what the model is thinking and each step it takes, and reopen it afterwards from “Worked for…”.
- The timer while a reply is being written now counts the whole time instead of starting over, and the status no longer sits inside an empty chat bubble.
- More credits on paid plans: Pro now has 10,000 credits a month (was 4,000) and Student 5,000 (was 2,000). The Pro trial has 1,000.
- Redo: write the latest reply again with one tap, with the same model or a different one.
- If the model you picked can't answer, another model answers instead, and the reply says which one.
- If nothing can answer, you get a plain explanation and a Redo button instead of a technical error.
Graphs, diagrams, running code and study tools
- Run Python: Apier can analyse a spreadsheet or CSV you attach, run statistics and simulations, test code by actually running it, and hand back the charts and files it makes. Free includes 5 runs a day; Student 40 and Pro 100.
- Flashcards and quizzes: ask to study or be tested and get interactive cards and multiple-choice quizzes with explanations and a score. Quiz scores are saved, so you can see your best, your last and your progress each time you retake one.
- Geometry figures drawn to scale, with labelled sides and marked angles.
- Molecules: the structure, formula and molar mass of any compound from PubChem, and structures PubChem doesn't have drawn from their SMILES.
- Citations in APA, MLA or Chicago from a DOI, ISBN or link, using the source's real details.
- Maps: places and routes on a real map for geography, history and travel.
- Team mode can now use all of these tools.
- Graphs are drawn from the math itself: ask for sin(x), a parabola or your data and get an exact, interactive chart in the reply (tap it to read values), with a PNG download.
- Bar, pie, line, scatter and area charts, with x axes in π or degrees for trig.
- Diagrams in the reply: flowcharts, cycles, timelines, mind maps, and sequence and class diagrams.
- Exact math: arithmetic, unit conversions, matrices, statistics, derivatives, integrals and solving equations are worked out by a calculator, not guessed.
- Chemistry and biology helpers: molar mass, balanced equations, and DNA tools (complement, transcription, translation, open reading frames).
- Essay stats: word count, reading level and overused words.
- The live status now stays in view under the reply as it's written, and shows what's happening between steps.
New models, readable math, and replies that show their work
- Claude Fable 5.2 replaces Fable 5.1: Anthropic’s strongest model, now 20% cheaper to use. Pick it from the model menu.
- Gemini 4 Carbon replaces Gemini 3.1 Pro as Google’s strongest model; Auto sends it hard science and math.
- Two new everyday models: Gemini 4 Hydrogen, which Auto now uses for medium-difficulty science and biology, and GPT-6.1 Terra, half the price of GPT-6.1 Sol, to pick yourself for cheaper everyday answers.
- Math looks like math: equations are typeset and centered, step-by-step answers have clear steps and a boxed result, whichever model answers.
- PDFs are read page by page, so worksheets, scans, diagrams and handwriting come through, not just the text.
- Live status while a reply is written: which model is working and what it’s doing, with a timer.
- Hard questions and long files no longer come back empty: replies get more room, and say so if they still run out.
- New file types: PowerPoint slides and PNG/JPG charts. Excel keeps numbers as numbers, and PDFs handle any language.
- Usage shows how many credits Auto saved you compared with always using the strongest model.
Smarter Auto, yearly plans, and student verification
- Auto reads messages better: it now recognises 14 fields, from code and math to law and creative writing, and spots when a message needs creativity, reasoning or a clear explanation.
- Model choices rebuilt from benchmark research: hard code and long writing go to Claude Opus 5.5, hard science and research to Gemini 3.1 Pro, everyday work to Claude Sonnet 5.5.
- New page: How Auto works, with every score, test and miss.
- Yearly billing for Student and Pro, at three months free.
- Student plan on a personal account: verify a separate .edu email with a 6-digit code.
- Terms of Service and Privacy Policy published.
Images, video, settings and plan limits
- Make images (including animated GIFs) and videos with sound, right in chat.
- Real PDF, Word and Excel files from a single request.
- Settings: theme, default model, reply language, personalization, location and time zone.
- Replies to a specific message, several chats replying at once, and “add to memory” from any message.
- Skills: saved instructions you can reuse.
- Clearer Usage page, with credits by model and a 5-hour pace so a month can’t vanish in one sitting.
Live search, file memory and a smarter memory graph
- Live web search with each provider’s own search tool.
- Attach PDFs, Word files, code and photos; later chats find them again when you mention them.
- Memory graph becomes a web: schools, jobs and projects become hubs, and corrected facts drop out.
- Export everything as one file; import memory from other assistants.
- A new look for sign-in and chat, matching the landing page.
First public version
- One chat for Claude, GPT and Gemini, with Auto choosing a model per message.
- Memory shared across every model, with a Memory window to see, edit and delete it.
- Team mode: specialists on different models work in parallel on big tasks.
- Free, Student and Pro plans with monthly credits.
Be first to try what’s new.
Apier is invite-only for now. Join the waitlist and we’ll let you know when your spot opens.
Join the waitlist