Sometimes the question is in a picture: a photo of a paper invoice, a screenshot of an email, a before-and-after of a job, a quote a supplier sent as a PDF. Attach it to your message the way you would in any chat app and ask. The photo or file is pinned to that message, shows inside your bubble, and the assistant answers with it in view - and keeps it in view for the questions that follow in the same conversation.
| Attachment | What the assistant does with it | Limits |
|---|---|---|
| Photo or screenshot (PNG, JPEG, WEBP, GIF) | Looks at the picture: describes it, reads the text and the numbers in it, compares two photos, or uses it as the reference for a new image. | Up to 20 MB each. A large photo is made smaller on the server before the assistant sees it. HEIC photos from an iPhone are not converted yet - share them as JPEG or PNG first. |
| Reads the document, text and layout, and answers from it: "what does the quote include?", "summarize page 2". | Up to 32 MB and 100 pages. | |
| CSV, Excel (.xlsx), text (TXT, TSV, MD) | Reads the content as text - a long table is shown to it as a typed table. Good for a quick look, a summary or a question about a handful of rows. | CSV and text up to 1 MB, Excel up to 5 MB; a table longer than 300 rows is shown in part. For exact totals over a whole sheet, page citations and a file you will ask about again, add it to a project instead. |
Up to 10 attachments per message. Attachments count toward your plan's storage (the same pool as Files) and are removed when you delete the message or the conversation.
Click + in the composer to pick photos or files, drop a file anywhere on the chat, or paste a screenshot with Cmd+V / Ctrl+V. While you drag a file over the conversation - the whole thread, the side panel, or the welcome screen - a card that says Drop to add covers the chat; let go anywhere and the file lands on the strip above the box. On the phone, the + button in the composer opens your photo library or the camera, and the file picker on builds that carry it. The + only adds photos and files; type / in the box to open the shortcuts menu. A file that is already in your organization's Files does not need to be downloaded first: on a new chat, the Files button under the box lists your newest files - search by name and click one, and a copy is attached to this message, copied directly on Genaya's own servers rather than through your browser, which is faster for a big file; Browse all opens the full Files picker, where you can pick several.
Each attachment shows above the box with a progress ring while it uploads and an x to take it off. A file that cannot be attached says why in plain words - too large, too many pages, or a format the assistant does not read.
"What is the total on this invoice?", "Describe what is wrong in this photo", "Summarize page 2 of the quote", "Which of these two quotes is cheaper?". The photo or file renders inside your message - click a thumbnail to see it full size - and the assistant answers with it in view. It stays in the conversation, so the next question can build on it without attaching it again.
Attachments land fast: the tile shows at once, a large photo is made smaller in your browser before it uploads, and several files go up side by side, so each one is ready on its own.
Attach a photo of a paper invoice, a screenshot of an email or a picture of a filled-in form to a chat message and ask Genaya to create the invoice, client, appointment or lead from it. It proposes the record with only what it can read in the picture, the same rule as pasted text: nothing is guessed, and money, status, recording and consent switches are left for you to confirm on the card. Nothing is saved until you confirm. The smart-paste pill on a create page reads pasted text, not a photo - for a photo, ask in the chat.
There is also a page made for it: open Genaya AI and press Creative in the left menu. Type what you want in the box, or press a cover under Templates (Logo, Flyer, Business card, Headshot, Product photo, Before and after, Storefront sign, Vehicle wrap, Price list and more), Trending (Brand mascot, Action figure, Magazine cover, Pop art portrait, Pet portrait, Blueprint and more), Social (Announcement, Quote card, Story, Carousel cover, Testimonial, We are hiring and more) or UGC (pictures that look like a customer or a team member took them on a phone: Selfie review, Unboxing, In hand, Flat lay, Car testimonial and more - the look that works in ads and campaigns). A cover never makes a picture by itself - it writes a ready request in the box that already describes the look of that cover, the page scrolls up to it, and you change the words if you want, add your photo with + when the idea works from one (a line under the box reminds you), and send. Genaya opens a chat and makes the picture right away in the look of the cover you picked, using what it already knows about your business - then tell it what to change ("make the title Growth Weekly", "use our blue"). It never invents a fact: a testimonial or a price list uses only the reviews, services and prices in your account, and it tells you when it finds none. All of it runs under the same plan and trial allowance as any other image. The same page makes short video clips too - see "How to create a video with Genaya AI".
Ask for a picture and Genaya AI makes one: "Create a before-and-after graphic for a chimney cleaning", "Design a square social post announcing our new clinic hours", "Sketch a logo in blue and white". Attach a photo and say "make one like this" or "redraw this on a white background" and the photo is the reference. Ask for HD when you need a sharper image and say square, landscape or portrait for the shape; one request can return up to four images.
Image creation is included on paid plans, with five standard images to try during the trial. If your workspace is not set up for it yet, the assistant says so in one line instead of making an image. A standard image is 20 credits, an HD image 60 (paid plans only), an exact-match edit that keeps faces and logos pixel-exact 40, each reference photo beyond the first 10, and a second round of drafts without a final image 10; the first round of drafts is free, and a refused or failed request spends nothing. A video clip on the same page is 30 credits per second of finished clip - 120 for 4 seconds, 240 for 8, 360 for 12 - on paid plans only, never on the trial.
Our team typically replies within one business day.