Vani Studio is open, plus Docs, Slides and Designs, Learning mode and scheduled tasks

Here's everything that's new since our last post. First, the highlights:

• Vani Studio: ready-made camera moves, effects, looks and photo apps for your images and videos, plus characters you can reuse.
• Docs, Slides and Designs: Vani writes documents, slide decks and designs you can edit, present, export and share.
• Richer replies: charts, diagrams, formulas and clickable choices, right in the chat.
• Learning mode: Vani tutors you with questions and hints instead of handing you the answer.
• Scheduled tasks: Vani runs a task for you every day, week or month and lets you know when it's done.

Vani Studio

Open Studio in the sidebar.

• Image: looks such as Portra 400 and Film Noir, and photo apps like Pro Headshot, Outfit Swap, New Background, Relight and Restore Photo.
• Video and effects: 30 camera moves you can mix (up to three per shot), 15 effects that bring a photo to life, and 8 transitions between two photos.
• Cinema (Pro and above): choose the camera, lens, colour grade and genre of your shot.
• Characters: save a person from 1 to 8 photos (you, or someone who agreed), then type @name to bring them into new images and videos.

Every image and video you make with Vani is now marked as made with AI in the file's metadata, which other apps can read. It doesn't change how it looks.

Docs, Slides and Designs

• Pick Document, Slides or Design in the + menu (or type /doc, /slides or /design), or just ask. Vani writes it in a panel next to the chat.
• Edit it yourself or ask Vani for changes, restore earlier versions, present your slides, and export to PDF, Word, PowerPoint or Markdown.
• Share makes a read-only link. Everything you make is under Documents in the sidebar.

Richer replies

Vani can now draw charts, diagrams and formulas in its replies, and ask you to pick from options with a click. Charts can be shown as a table, copied as CSV or downloaded as an image.

Learning mode

Switch on Learning in the model menu. Vani asks what you already know, gives one hint at a time and checks your understanding, and still shows the full solution when you ask. Make it the default for new chats in Settings → Personalization → Defaults.

Scheduled tasks

Choose Schedule a task in the + menu, or write something like "every weekday at 9am, summarise my open Linear issues" and Vani offers to schedule it. Tasks can search the web, run research and read your connectors, and results arrive in Updates. Manage them under Tasks in the sidebar. Free includes one daily task.

Projects, skills and connectors

• Project knowledge: add PDF, Word, text, CSV or code files under Project knowledge on a project's page. Every chat in the project can use them, and replies show when they do.
• Community skills: share your skills with Publish to community, and find other people's in Skills → Community. Our team reviews every skill before it's listed.
• Connectors directory: browse connectors under Connectors in the sidebar. Paste a Notion, Linear or Sentry link and Vani offers to connect it.

Also new

• Message actions: hover over a message (or tap it on a phone) to see when it was sent and to Copy, Edit, Retry or Fork it.
• Less usage on long chats: with Claude Sonnet 5.5, Opus 5.5 and Fable 5.1, Kimi K2.8 Preview and MiMo V2.6, the part of a chat the model reuses from its cache now uses a fraction of the usual allowance.
• Monthly recap: your month on Vani in Settings → Recap, made from counts and chat titles, never your messages.
• A quick catch-up: if you've used Vani before, you'll see a one-time look at what's new and a review of your settings. Nothing changes unless you save.
• More in your export: Settings → Your data now also includes your memories, skills, documents, project knowledge, tasks and Studio characters.
• Models: compare every model side by side at vani.ai/models.

For developers

• Every image and video model from chat is now on the API: GPT Image 2, GPT Image 2.5 Flare and Sunburst, Grok Imagine Image and Grok Imagine Video join the others, priced per image or per video (see pricing).
• Cached tokens in usage: prompt_tokens_details.cached_tokens (and its Responses and Messages equivalents) shows what was billed at the cached-input price, and Claude cache_control hints are passed on where supported.
• Context sizes: GET /v1/models now lists each model's context_length and top_provider.max_completion_tokens.
• Learning mode: add "vani_mode": "learning" to a Chat Completions, Responses or Messages request.
• Guides for Xcode, VS Code, JetBrains and the OpenAI and Anthropic SDKs are at vani.ai/docs/guides, and recent Codex CLI versions work with /v1/responses again.

Fixes and improvements

• Replies from our support team now show in Updates.
• Images from other websites in a reply wait until you click Load image.
• The sidebar, your plan and the usage gauge update right away, including changes from your other tabs.

Coming soon

• Code & files: Vani runs code on your files in a private sandbox, to analyse data and make charts.
• Browser: Vani opens and reads web pages for you.
• One-click deploy for apps you build with Vani.
• More models for images, video, lip-sync and music.

We're still fixing bugs. If something doesn't look right, please use Report a bug in the app; we read every report.