Changelog

Every change that reached the site, newest first. New tools, fixes you would notice and languages added, each with a link so you can try it.

RSS feed
  1. Font pages: Find similar fonts now searches with this font's own sample, plus text fixes

    On a font page, the Find similar fonts from an image button now uses the sample of the font you are viewing, so you see similar fonts right away without picking an image. The image is made in your browser and is not sent to our servers. We also fixed font descriptions that left a blank where the maker's name should be, and a doubled period at the end of some license text. For the Japanese pixel font でんきチップ, the license is now shown as SIL OFL 1.1 as in the official release, the inverted sample image is fixed, and you can try it right away from our font server.

    Font finder: fixed "Couldn't load font data" when searching with a Korean image

    After yesterday's 15 new fonts were added to the list, searching with an image of Korean text showed "Couldn't load font data" instead of results. The new fonts do not have search-by-image letter data yet, and the search stopped because it expected every font to have it. Fonts without that data are now skipped, so Korean search works again. If a search fails in the future, the error message now ends with a short reason so we can fix it quickly.

  2. Font finder: a "Newly added" section at the top, plus 15 recently released Korean fonts

    The font list now opens with a section for newly added fonts. Each time new fonts arrive, the latest batch appears there and the previous batch moves down into the main list. This time we added 15 recently released Korean fonts. Hakgyoansim Edge, Deongdeongi and Bamhaneul, HEEJOY, Gyun Gothic, Gyun Myeongjo, BeyondVillage, Yar, Shinyar, Gamdasal, Modu Dunggeul Pen, BSM Smile Lasik, BSM Vision Correction and HS Hwalgong Gulim allow redistribution, so you can type your own text in them right away. REATIC does not allow redistribution or web embedding, so it is shown as a sample image with a link to the maker. Every font was checked against its maker's official page, and its license is listed. They will appear in search-by-image results with the next data update. Also, the Find similar fonts from an image button on font pages used to just open the font finder; it now lets you pick an image right away and shows the matching fonts.

    Font finder: license file missing from some Google Fonts ZIPs, and the closest font card now follows the filters

    The Download all weights ZIP on Google Fonts pages includes the license file, but a few fonts whose Google Fonts source folder has no license file (M PLUS Rounded 1c, Tinos, Kumar One Outline, the Edu families and others) were missing it. We checked all 1,583 Google Fonts families. For these fonts the ZIP now gets an OFL.txt made from the copyright line in the source data and the full OFL 1.1 text, and Ubuntu Sans, whose file is named LICENCE.txt, now gets its license too. Also, after finding fonts from an image, changing the price filters (paid, included with Mac and Windows) updated the list but left the closest font card showing the top match from before the filter. The card now shows the top font of the filtered list. Filter labels that were long enough to wrap onto a second line are now shorter, so each fits on one line in every language.

    Font finder: try 740 more fonts with your own text

    Fonts outside Google Fonts and noonnu could only be seen as sample images. We checked each license again. The 646 fonts that allow redistribution are now on our font server, so you can type your own text in them right away in the list and on the detail page, just like Google Fonts. The detail page shows the original source, the license and a link to the full license text. For 94 fonts that cannot be redistributed but whose makers allow it from their own servers, press Try it and your browser downloads the font straight from the maker and uses it only in that browser. Nothing passes through our servers, nothing loads until you press the button, and the size is shown on the button when it is over 5 MB. Other fonts, and fonts whose web use is paid, still show a sample image and a link to the maker.

    Photo reports: an Options button on the table forms, with before and after only and 8 photos per page

    The site photo form and the 6 or 8 per page form split photos into before, during and after, and the 6 or 8 per page form had no way to choose 8 despite its name. Many offices only ask for before and after, so both form cards now have an Options button in the top right corner. It lets you choose the photo stages (before, during, after, or before and after only), and on the 6 or 8 per page form also the number of photos per page (6 or 8). With before and after only, the site photo form holds two larger photos a page, and the 6 or 8 per page form pairs each row with before on the left and after on the right. When you add more photos, the pages continue in the same layout. It works the same in Excel, Hangul and Word, and your choices are remembered for next time.

    Font pages: weights now follow the official release, and image-only fonts describe what the page really does

    Detail pages for fonts that are not on Google Fonts (Pretendard, Paperlogy, Gmarket Sans, S-Core Dream and others) showed the wrong weights. Pretendard, which has 9, appeared as "2 weights" and "400, 700", because the page listed the two weights used for font matching. The site now downloads the official release, counts the weights it really contains, and shows the count and names (Thin to Black): 9 for Pretendard, Paperlogy, S-Core Dream and SUIT, 3 for Gmarket Sans. Fonts that could not be checked against official files hide the weight information instead of showing a wrong number. Link previews for fonts shown only as sample images also said you could preview your own text; they now mention only what the page offers: a sample image, the license and where to download it officially.

    PDF Editor: password-protected PDFs now open, and restricted PDFs no longer save as blank pages

    Opening a PDF with an open password only showed "This PDF could not be opened" and went back to the start screen, so it looked like the file was broken. The editor now asks for the password and opens the file for editing once it is correct. A wrong password lets you try again, and Cancel goes back to the start screen. When you save, the same password is added back by default; turn it off in the dialog to save without a password. The password is used only in this browser and is never sent anywhere. PDFs that only restricted printing or copying did open, but the saved file came out as blank pages, because the encrypted content was written out without being decrypted. The editor now decrypts them first and puts the same print and copy restrictions back when you save.

  3. PDF Editor: editing a line with parentheses no longer leaves the old text behind

    In PDFs made with the standard fonts, editing a line that contained parentheses, such as "(books included)", failed to find the command that drew the original text, so the new text was drawn on top of the old one. The saved file covered it with a background box, so it looked fixed, but the old text stayed in the text layer and showed up when copied or searched. The cause was comparing the text without undoing the way PDF stores parentheses and backslashes (written as \( and so on). The editor now finds those lines and changes them in place. In the rare cases where the original text still cannot be changed in place, it no longer covers it up when saving. The edit is withdrawn and the editor tells you the line was left unchanged, so nothing looks edited while the old information remains.

  4. New tool, AI Detector: check whether a file was made by AI or a script from the traces inside it, not the writing

    Drop a file and get a one-line result: an AI generation label was found, the file was made by a script, a copied AI answer left traces, or no traces were found. Below it, each line shows which field said what. For documents it looks for what python-docx leaves when ChatGPT builds a Word file (the author, the fixed 2013 date and the description) and the defaults of python-pptx, openpyxl, PptxGenJS and ReportLab, each confirmed by building a file ourselves. For images, video and audio it reads C2PA Content Credentials (JPG, PNG, WebP, AVIF, HEIC, MP4, MOV, WAV, MP3, and inside Word and PDF files), the IPTC AI label and the new IPTC 2025.1 AI fields, and Stable Diffusion generation data, and it checks C2PA signatures in your browser with the official C2PA library and the C2PA trust list. Links copied from ChatGPT answers (utm_source=chatgpt.com) and citation markers are caught too. It gives no probability and only reports what the file says, and it states that pasted text and invisible watermarks such as SynthID cannot be checked. View file metadata now uses the same check, flags any traces at the top, and shows the document description field.

    New Hangul document engine (rhwp 0.9.0), and correct text size and weight in PDF to HWP

    The HWP Editor, the Hangul converters, previews and the photo ledger now run on a new version of the Hangul document engine. It fixes HWP files with an empty header or footer that Hancom Office would not open after saving, body text that could go missing after a save, an error when saving curved shapes to HWPX, inserting, moving and undoing pictures in table cells, and table and text layout. HWP to PDF now lines up text and titles closer to Hangul, and stepped gradients carry over without making the PDF large. Testing also turned up an old problem in PDF to HWP and HWPX: text was saved with a size of 0, and the bold, larger heading style spread to the whole body. Files now start from a Hangul blank document, with 10 pt body text and 16 pt bold headings. The engine is downloaded once, about 10 MB.

    PSD Editor: template text no longer jumps off the canvas when edited, and large files show progress

    Some designer templates keep a text layer anchor far off the canvas while the text is drawn somewhere else, so editing or moving it with the Type tool threw the text off the canvas. Such text is now placed where Photoshop drew it when the file opens, so it stays put as you edit; untouched text is saved with its original data. Large files that take a minute or two to open, such as a PSD over 90 MB, now show the file name, size, a progress bar and an estimated time instead of a blank screen, and if the editor runs out of memory and stops, it says so and offers a Restart button. A file opened from the site opens again by itself after the restart.

    PSD Editor: the Type tool no longer picks text in hidden groups

    Designer PSDs often keep alternative versions in hidden groups. Clicking visible text with the Type tool could pick a text layer from a hidden group sitting in the same place, so the invisible text changed instead. Hidden layers and layers inside hidden groups are now skipped, and when several overlap the topmost visible one is picked, as in Photoshop. Auto-Select with the Move tool follows the same rule. Saving to PSD also marks shape layers the way Photoshop does, so other apps read them as shape layers rather than solid color fills.

    Photo reports: pick up where you left off, make it like last time, my layouts

    The new photo report screen now has three shortcuts on its first page. While you work, the document and photos are kept in this browser, so "Pick up where you left off" opens them again as they were. "Make it like last time" starts with the same form and file type, and if you choose photos or drop them on the first page they fill in as soon as it opens. After you mark the photo cells of your own Excel, Hangul or Word form it goes into "My layouts" and opens in one click next time, and the layouts and reports saved on the previous screen show up there too. Photos no longer go missing when you save a report made in Word.

  5. Photo reports: make them right in the Excel, Hangul, Word and PDF editors

    The photo report maker has a new screen. Pick a form and it opens straight in the Excel, Hangul or Word editor, with a photo library always open on the left. Add all your photos at once and press auto fill to put them into the photo cells in order, adding pages when the cells run out. Drag a photo into any cell, or drag one photo in the document onto another to swap them. Change tables and text freely in the editor. Designed reports such as construction and cleaning reports open in the PDF editor, with the photo cover and checklist. You can also open your own Excel, Hangul or Word form and mark its photo cells once to fill it the same way. On phones the previous screen still opens.

    Photo reports: edit cells before adding photos

    Before you add any photos, you can now click the caption, location, work notes, before/during/after and date cells on the empty document and edit them. When you add photos later, what you wrote moves into those photo cells. Dates are filled in from the photo information first, and you can click to change them. Edits to the title of an inspection report now stay too.

    Photo reports: one click to edit any text with a pencil mark

    On a computer too, a single click on text with a pencil mark lets you edit it right away. Clicking another cell saves what you typed and moves on to edit that cell.

  6. JTD editor: double box borders, text colours and margins

    Double borders round text boxes are drawn, and each text box shows its own text colour, such as navy or green headings. Text box margins are read in the right order, top and bottom then left and right, so the text starts where Ichitaro Viewer puts it.

    JTD editor: pictures in mid paragraph and shape outlines

    A picture attached in the middle of a paragraph now sits beside the line that holds its character, so pictures that covered the text appear where Ichitaro Viewer puts them. Shape outlines take their own colour, so the black edges of holly and a candy cane show, and striped fills, very thin lines and thin text box borders are drawn.

    JTD editor: table colours and cell alignment, text box spacing, wood grain title boards

    Ruled line colours, cell shading and the alignment set on each cell (days at the bottom, weekdays centred) now match Ichitaro Viewer. Text boxes follow the line spacing and characters per line saved in the document, so their lines break in the same places. Shapes filled with a texture such as wood grain, shapes without an outline and pictures attached to table rows appear where they belong, and empty lines pushed down by text boxes now move the same way, so calendars land in their original place.

    Photo Report Maker: change dates, stages and labels right on the page

    Tap a photo date to pick another day, or change several photos at once. Changed dates go into HWP, PDF, Word and Excel files, and the preview marks dates that differ from the photo with a small dot. Photos without a stage show a "Pick a stage" spot instead of an empty bold line. Row names and order beside the photo, extra header fields and photo width can be saved to My layouts and opened from the first screen. Photos whose time is written with a single digit hour now get their date too.

    JTD editor: headers, footers, page numbers and pixel pictures

    Headers and footers saved in a document now appear in the same place and alignment as in Ichitaro Viewer, with the page number filled in on each page. Low resolution pixel pictures keep their dots instead of turning blurry.

    JTD editor: drawings that went missing or into the wrong place

    Drawings are now matched to their places in the order the document gives rather than by size, so drawings that went missing or into the wrong place now appear where they belong. Curved drawings are measured by the curve itself and no longer look squashed.

    JTD editor: drawings, page backgrounds and frames in school newsletters

    Flowers, carp streamers, people and lines drawn in school newsletters now appear in colour as in the original. Page background colours and background pictures, such as the gold border of a certificate, are shown, and text box lines, rounded corners and stacking order follow Ichitaro Viewer. When a large frame surrounds the whole page, the text stays inside it.

    JTD editor: school newsletters written across, vertical text boxes and certificates

    School newsletters written across that opened as vertical documents now open as they are, so calendar-style event tables lie across the page. Vertical text boxes on such pages are drawn vertically, and documents that give only the paper width, such as an A3 landscape certificate, open at the right size. Body boxes anchored on the last empty lines, which went missing, now show.

    JTD editor: three columns and title frames in vertical newsletters

    Vertical school newsletters split into three bands down the page now show those three bands as in the original, with the same number of characters per line as Ichitaro Viewer. Ruled frames round titles are drawn, and photos, text boxes and drawings in the text sit where the character they belong to is.

    JTD editor: text boxes and columns in vertical newsletters

    In vertically written documents each text box now keeps its own direction, so a school newsletter shows its body and title top to bottom and its heading boxes across, as in the original. Text boxes also keep their character size, so large titles appear at full size, and boxes and figures sit where they are on the paper. Columns in vertical documents are split top and bottom on the page.

    JTD editor: line counts, vertical writing, columns and two pages per sheet in school documents

    School documents set to 43 to 70 lines per page no longer spread out to 40, so each page holds what the original does. School newsletters written vertically that opened across now open vertically. Two-column documents fit the same number of characters per line as the original, and grade newsletters printed two pages to an A4 landscape sheet show as two pages, as in Ichitaro Viewer.

    JTD editor: school handouts laid out almost exactly like Ichitaro Viewer

    Text now keeps clear of pictures and text boxes fixed on the page from the first line, so text beside labels and around route maps no longer overlaps. Empty lines and drawings under title pictures land in the right place, and runs of punctuation such as 。( take half a cell so lines fit as in the original.

    Your last settings come back, and snapshots to the pixel

    Image convert, compress and resize, PDF compress, PDF to JPG and PNG, QR codes, photo reports, the mail viewer, CAD PDF export, chat backup export and the HWP editor’s save-as menu now start with the format and settings you last finished with. Reset to default is always one click away, and settings stay in this browser. In the PDF editor, Snapshot shows a magnifier with pixel coordinates while you drag, snaps to text edges, moves 1 px with the arrow keys and takes the area in px. The Snapshot screen now speaks 20 languages.

    PDF editor: Snapshot, and your last format comes first

    Snapshot in the toolbar (Alt+S) draws the chosen area again from the PDF, up to 600 dpi. Adjust brightness, contrast, grayscale, crisp document, transparent white, rotation and marks, then send it to the clipboard, a PNG, JPG or WebP file, a new page, this PDF, another PDF tab or the printer. Table text can be copied too, with cells split by tabs. PDF conversion on the start screen now shows format chips with the one you used last in front, and tools you used more than once gather under Your tools.

    JTD editor: B4 handouts split into top and bottom halves

    B4 landscape handouts laid out in top and bottom halves now flow like Ichitaro Viewer: top left, top right, bottom left, bottom right. Text boxes and pictures in the lower half land in place, empty lines no longer slide down round pictures, and lines with ruby stay one line tall.

    JTD editor: text box frames and shadows, handouts laid out like the original

    Text boxes now show their dotted or solid frame lines, rounded corners, shadows and solid or gradient fills. Dotted fills of figures and pictures placed inside text boxes show too, and body text flows around photos. A school handout was matched line by line against Ichitaro Viewer. Saving as Word now keeps figures.

    JTD editor: figures and art text now show

    Figures drawn in Ichitaro documents (illustrations, balloons, ellipses and rounded frames, staves) and art text and drawings made in other programs such as Hanako now appear in the document. Two-column layouts such as school handouts now fit their columns and come out close to Ichitaro Viewer.

    JTD editor: the document's own fonts, text boxes, columns, pictures in Word

    The font list now shows the font names the document uses, and the one you pick is saved to .jtd. Rounded Gothic, textbook, brush and pop faces are drawn with similar free fonts. Text inside text boxes and documents set in two or three columns now show in place, body text flows around text boxes and photos, and forms with tight line spacing get the same line height as Ichitaro. Saving as Word now keeps the pictures.

    JTD editor: Delete works while typing Korean, and a tidier tool palette

    With a Korean input method on, Backspace, Delete and the arrow keys did nothing and syllables fell apart into letters. That is fixed. The tool palette on the right now keeps each label and its controls on one line, with bold, alignment and row buttons as icon groups, so nothing wraps onto the next line.

    JTD editor: vertical documents stay vertical

    Ichitaro documents written vertically now show and edit vertically. Kanji and kana stand upright, 、。 sit at the top right of their cell, long vowels and brackets turn with the line, and ruby sits to the right. Arrow keys and the cursor follow the vertical direction, and Word export keeps the text vertical. Small stray marks next to pictures are gone too.

  7. JTD editor: pictures in documents now show

    Photos and illustrations placed in Ichitaro documents (GIF, JPEG, PNG, BMP) now appear where they sit on the page. Their positions were checked side by side with Ichitaro Viewer, and they are included in PDF export and the page thumbnails. Pictures stay in place when you edit the text and save as .jtd.

    JTD editor: compressed Ichitaro documents (.jtdc) open too

    Besides compressed templates (.jttc), compressed documents (.jtdc) can now be picked in the open dialog. All 17 real public files we tried opened, and their text matched Ichitaro Viewer at 99.2%. After editing they save as .jtd, and large documents of around 100 pages also save as .jtd and open in Ichitaro Viewer.

    JTD editor: choose Mincho or Gothic

    The font box in the format bar now switches between Mincho and Gothic. The choice is kept when saving as an Ichitaro document (.jtd), and the saved files open in Ichitaro Viewer.

    JTD editor: compressed Ichitaro files, help when a file does not open

    Compressed Ichitaro documents (.jttc) now open. When a file cannot be opened, the editor says what the file is, why, and what to do, and Word or PDF files are handed to the right tool. RTF or text files with a .jtd name open as a new document with their text. Gothic text shows in a Gothic face and stays Gothic in Word export. Escape in find no longer opens the menu, and dragging a selection past the edge scrolls the page. Documents with a page break just before a ruled box now save as .jtd.

    Photo Report Maker: cover titles break the same in preview and files

    Cover title line breaks and size are now decided when the document is built, so PDF and HWP files match the preview. Past three lines the title shrinks instead of being cut. Japanese titles without spaces are split into even lines, preferring breaks after の, + or ・ and before brackets. The cover starts with one photo slot per result photo, and added slots are picked by tapping them.

    Photo Report Maker: SVG logos, cover photo slots, addressee and wrapping

    SVG logos are now converted and used. An image that cannot be read is reported and the previous logo is removed so an old logo never ends up on a document, and white logos sit on a brand color plate. The cover makes only as many photo slots as there are after photos, adjustable from one to three. You can pick the addressee ending, and long cover titles, taglines, fire safety work notes and site locations wrap before anything is cut.

  8. Photo Report Maker: before and after headings on fire safety logs

    Each item on fire safety submission logs now always shows "Before" and "After" above its two photos, so long captions no longer push the stage out. Trade header fields such as the building name, address and inspection date sit right under the company details instead of being folded away.

    Photo Report Maker: fire safety pairs, long site names, painting steps, preview pages

    Fire safety logs pair before and after photos by matching descriptions, or else by the after photo right after it in file order, instead of lining them up separately. Words like 退去時 are removed from restoration captions. Long project names and locations in submission headers shrink or wrap to two lines instead of being cut. Painting reports keep the stage column to before, in progress and complete, put the process name in the description and keep file number order within a stage. Periods taken from photo dates are marked as such. The preview now has the same page count as the saved file, and empty text fields are added from the side panel.

    Photo Report Maker: ZIP name fixes, stage labels and a repair report

    Korean file names inside ZIPs no longer turn into Thai letters when they contain symbols such as a middle dot, and ZIPs marked as UTF-8 are read as UTF-8. Stage words written with a space, such as "after work", are now read as the stage and removed from the caption, and file name stages win over shooting dates. Without after photos the cover uses good inspection photos, then any photo, and changing one cover photo keeps the others in place. The chosen document name now matches on the cover, the file name and the summary, and a repair report is back under More.

    Photo Report Maker: company cover by default

    The first page of client reports now looks like an official company document: a brand color band, your logo and company name, a tagline or contact line, the document type and title in the middle, and if you like three result photos with a one line note each. The brand color is picked from your logo and can be changed, and it is remembered with the tagline for next time. Cover title size comes in small, medium and large. With the photo cover, auto now only uses after photos and falls back to the company cover when there are none. PDF, HWP, Word and Excel all get the same cover. The document type stays above a custom title, start and end dates you type go into the cover period, and you can tap the cover photos in the preview to swap them.

    XLSX editor: opens web page tables saved as .xls

    Many admin systems export a web page table (HTML) and just name it .xls, so other apps refuse it as HTML. The XLSX editor now pulls the table out and opens it as a sheet, reading old encodings such as EUC-KR, turning spanned cells into merges, keeping numbers as numbers and leading zeros intact. Saving gives you a real .xlsx file.

  9. Photo Report Maker: smoother editing on the page

    On fire safety logs the work note you type now goes under both the before and after photo, and dragging a photo to swap no longer shuffles the others. Fields you can fill on the page are gone from the side panel, which keeps only company name, logo, cover photo and the cleaning checklist. Restoration reports and move-in/move-out records are now separate choices, document names match the document you picked, and photos without names or dates no longer repeat the header row for every photo. A photo added to an empty slot of a fire safety item now joins that item.

    Photo Report Maker: pick the document first, then drag photos in

    The first screen now asks what you are making, such as a fire department photo sheet, a client job report or a cleaning completion report, and opens an empty document whose slots say what goes where. Drop photos and they fill the slots in the order they were taken. Drag a photo onto another to swap them, or on a phone tap a photo and then the slot. The separate checking step is gone, and empty slots and unplaced photos are listed at the top. Saving is one button in the format that suits the document, and next time you can start straight from the document you made last.

    Photo Report Maker: tuned with real site photos

    Sets split only by date now get before and after from the whole job period, and daily reports and unsorted photos get no step label. Folders with cleaning in the name pick the cleaning trade, and the cleaning checklist comes in move-in, office, common area and air conditioner versions. Without an after photo you pick the cover photo yourself, and daily reports can be saved as one file per day. The fire safety pairing screen shows smaller photos in two columns.

    Photo Report Maker: 8 per page, A4 landscape, daily reports, checklists

    Photo logs can now hold 8 photos per page or run in A4 landscape (3x2, 4x2), with labels on the photos if you like. New: a daily report with today’s work and tomorrow’s plan per date, a cleaning checklist of areas and tasks, and a routine visit report whose site details you can save and load next month. For fire safety logs you can swap before and after photos per item on the checking screen, and trade step labels are only set when the file name says so or you pick them.

    Trade presets in Photo Report Maker

    Pick fire safety, painting, leak, move-out or cleaning and only the boxes and step labels that trade needs appear. Fire safety gives a 6-photo before and after fix log with property, address and inspection date; painting adds step labels with materials, coverage and inspection; leak adds when it was found, where, cause and result; move-out adds damage and who pays for each spot with a summary. The cover can now show site, work period, contractor and dates, and you can add approval boxes, use Pass/Fail/N/A and pick a document name.

    Photo Report Maker client reports are easier to read in print

    Text is larger (captions 10pt, labels 8.5pt, footer at least 8pt). Every page carries the company, report name and page number at the top, and each set shows its dates, trade, photo numbers and status in one row. Photos are numbered 1-1, 1-2 and so on, matching the summary table. Photo sizes are chosen so pages do not end half empty, and move-in records put the condition boxes right under each photo with signatures on the last page.

    Photo Report Maker: multi-line captions and Word page count fixed

    Client reports now show every line of a caption, and you are told while typing if a caption is too long to fit. Word files no longer run to more pages than the PDF. Photos taken seconds apart are kept in the same area, and moving photos to another set shows each set with a thumbnail and photo count.

    More places to write in Photo Report Maker

    Each report now has its own text boxes, such as summary, notes, cause and action, or overall assessment, and every room can have a line on what was done there. Inspection reports let you pick Pass, Watch or Fail and an action for each photo, and move-in records let you mark Clean, Scratched or Damaged. Common phrases go in with one tap, and phrases you used before are suggested from this browser. Photos can be moved to another set by ticking or dragging them, and before and after cleaning photos taken in the morning and afternoon of the same day are paired in order. The duplicate warning now appears only for identical files.

    Photo Report Maker reads file names better

    Without folders, names like "livingroom_before", "01 mold - fix" or "01_kitchen_sink_fail" now sort photos by room, case and status. Move-in and move-out photos pair up by room, and inspection counts only the statuses written in names. Photos that cannot be opened are left out with a note, and photos without a capture date no longer get today’s date. Word files now have the same page count as the PDF, with fewer half-empty pages.

  10. New: Photo Report Maker

    Drop your site photos and they are sorted by folder or date taken, labeled before, during and after, and laid out as a photo report. Pick one of three photo log tables or six client reports such as renovation, cleaning and move-in inspection, and save as Word, Excel, PDF or HWP. Double-click a photo to rotate, crop or hide part of it, and double-click any text to change it. Save your layout to reuse it next time. Your photos never leave the browser.

    The Word, Excel and PowerPoint editors and the CAD Viewer now open looking like editors

    Before you pick a file you now see the menus, ribbon and status bar of the editor, and you open a file or start a new one in the middle. The Excel editor shows a faint sheet behind. The CAD Viewer opens in its DWG layout (toolbar, layer panel, sheet tabs) and switches to the right view for the file you open. The PDF and PSD editors are now named PDF Editor and PSD Editor.

  11. Go straight to the tool from any article

    Every article now has a button that takes you to its tool. It sits at the bottom center on phones and the bottom right on computers, and stays visible however far you scroll. Close it with X and it stays closed for that article.

    Every page opens much faster

    Every page used to download the interface text for all 20 languages at once. Now it only downloads the language you are reading, so the first screen loads about a fifth of the script it used to. Ads and visit statistics load after the page has finished drawing, and the font is now a file holding only the characters the site uses, so a Korean page downloads 190KB of font instead of 500KB, once for the whole site. If you picked a language before, the site now opens straight in that language without showing the English page first.

    CAD Viewer: choose the units of an STL or OBJ

    STL and OBJ files carry no units, so the viewer guesses from the size. You can now pick mm, cm, m or inch next to the unit line in the Model tab, and measurements and the sizes in Properties update right away. The unit guess no longer shows up as a "some parts may be missing" warning.

    Font finder adds 240 HY fonts and fonts it was missing

    We added 240 HY fonts from the official Hanyang I&C store (FontBada), and common document fonts such as HY HeadLine, HY Baekseong and HY Dongnyeok are now matched letter by letter. Fourteen Korean fonts confirmed free for commercial use by their official sources and 61 Google fonts that were missing, including Roboto, Open Sans and Noto Sans, are now in the list. The recognition model also learned from text over real photo backgrounds, and in a 300-photo test the right font now comes first in 41% of photos, up from 39%.

    New: Fix garbled CSV and Check PDF fonts

    Fix garbled CSV reads a CSV in the right encoding (UTF-8, Shift_JIS, CP949, Windows-1252 and more), shows it as a table, opens it in the sheet editor or saves it again as Excel-friendly UTF-8, the original encoding or .xlsx, with leading zeros and dates untouched. Check PDF fonts lists the fonts in a PDF, shows which are not embedded or copy as garbage, and can save a copy as images that looks the same everywhere. The sheet editor now opens CSV files directly, Extract ZIP opens .tar.gz, .tgz, .tar.xz and .tar.bz2 in one go, and XLSX / CSV / JSON no longer turns CSV dates into numbers or drops leading zeros.

    Drop the wrong kind of file and we send it to the right tool

    If you drop a file a tool can’t open, like a PDF in the CAD Viewer or an HWP in Extract ZIP, you now get a card such as "This file is PDF. Open it in the PDF Editor?" instead of an error. One click opens it there without choosing the file again. Password-protected PDFs go to Unlock PDF, scanned PDFs with no text get a short note on what you can do, and formats we don’t support yet (videos and so on) are named plainly. The file only moves inside your browser.

      English text left in the PSD editor is translated

      Effect names in the Layers panel (Effects, Stroke, Satin and the rest), the effect colour picker, and the status line and notices after opening or exporting a file now follow the editor language. The notes shown when a JPG export flattens layers or fills transparency with white are translated too.

      Measuring inside a sub-figure of an SXF P21 drawing gives the real length

      In P21 files written with AP202 entities only, a measurement inside a sub-figure showed the paper distance alone. It now also shows the real length at that sub-figure’s scale (1:100, 1:30 and so on), as it already did for SFC files.

    • The CAD viewer draws MLIT P21 (AP202) drawings the same as their SFC files

      SXF P21 files written with AP202 entities only used to show just the frame and title block. Scaled sub-figures, dimensions and leaders, Japanese text and layers now come out exactly as in the SFC file of the same drawing. When something still cannot be drawn, the status line names its kind.

      Rotate text, shapes, images and QR codes in the QR designer

      Drag the round handle under a selected element to turn it to any angle; hold Shift to snap in 15° steps. The side panel has a slider, 90° buttons and Straighten. Rotated elements come out the same in PNG, SVG and PDF.

      Font finder learned from real designs

      We retrained the recognition model with about 6,000 more lines of text from real designs and photos such as posters, book covers and video captions, and added per-weight data for paid fonts. In a 300-photo test, the right font now comes first in 39% of photos (up from 34%) and in the top five in 54% (up from 49%), and outlined or shadowed lettering is recognized more often.

    • MCP server 0.8 finds the font in a picture and fetches a missing free font

      The EZPZ File MCP server that AI agents call has a new tool, font_find, which makes twelve. It reads the text in a picture or on a page of a PDF, HWP, JTD or XDW and ranks the fonts it looks like, the same way as the font finder on this site. Each match shows how close it is, whether it is free, paid or bundled with the OS, its licence and, for free fonts, where to get it, and the picture never leaves the computer. When you put text into a PDF template such as a business card and its original font is not installed, the same free Google font from the font finder list (Nanum Myeongjo, for example) is fetched once by name and used.

      MCP server 0.7 edits AI and PSD files, makes many files from one list and reads CAD drawings

      The EZPZ File MCP server that AI agents call now handles Illustrator (AI, EPS, SVG) and Photoshop (PSD, PSB) files. It lists layers and objects by id, hides them, changes text, and draws a single layer on its own so the agent can find it by its shape. Give it a CSV or XLSX list and a design such as a business card becomes one file per row, and PDF designs are filled in their original font and stay small. CAD drawings such as DWG, DXF and JWW can be read for their text and turned into PDF, PNG or SVG. There are still eleven tools, and the design engines (about 13MB and 6MB) are fetched only once.

      PDF editor: edited digits and Latin letters no longer turn into boxes in LibreOffice PDFs

      When you edited text in a PDF made with LibreOffice, digits and Latin letters could show up as empty boxes (□). This is fixed, and the edited text now shows as typed.

      MCP server 0.6 edits HWP and Word files, fills templates and redacts PDFs

      The EZPZ File MCP server that AI agents call now has eleven tools, up from six. It looks inside an HWP, HWPX or Word file to find paragraphs, tables and fields, edits them by those addresses, and fills {{name}} templates into new documents. It redacts PDFs so the text is gone from the file, draws pages as images, and checks a file before it goes out for page count, empty spots left and words that must not appear.

      PDF to CAD: circles and arcs saved as CIRCLE and ARC, no more sheet-edge lines

      Circles and arcs in drawing PDFs (grid bubbles, door swings, north arrows) were split into hundreds of short lines, so they vanished in the full view and only showed after zooming in. They are now saved as CIRCLE and ARC, and connected lines as one polyline, so they show from the start. The white page background in PDFs printed from a web browser no longer leaves lines along the top and left edge of the sheet. The CAD Viewer also draws circles that other programs saved as short lines in the full view.

    • XDW Editor: crafted files are stopped, and covered text stays out of PDFs

      Sizes written inside a file are no longer trusted, and a file past the limits is refused with a notice: a few hundred bytes can no longer make the tab use gigabytes of memory. Files are checked in a separate worker before they open. Text hidden under a black box or a picture is left out of the text in PDF and text exports, and the editor tells you how many characters it left out. Engine: ezpzfile-xdw 0.3.0.

      JTD Editor: files with many style changes no longer use up memory

      Thanks to a second private report from HIDANO Masaki (KUKAN project), ezpzfile-jtd now shares character style data instead of copying it at every change and for every glyph. On a crafted file, memory went from 1.4 GB to about 40 MB. A file whose style changes go far beyond normal is refused with a notice. Engine: ezpzfile-jtd 0.3.3.

      PDF editor: files in tabs, an Organize view and permanent redaction

      Drop several PDFs and each opens in its own tab, keeping its edits and undo history as you switch. “Merge tabs” joins the open files in tab order, and dragging a page onto another tab copies it there. Organize shows every page on one screen to reorder, rotate, delete, extract, split and insert. Areas covered with Redact are removed for good when you save, so copying or search can’t bring them back (redacted pages are saved as pictures). The tool bar now sits at the top, and the start screen has quick tools for edit, merge, organize, redact, sign and split.

      Vector Editor: tool shortcuts work with a non-Latin keyboard layout

      Single-key tool shortcuts such as V, P and M now go by the key’s place on the keyboard, so they work while a Korean or other non-Latin input mode is on. While you type text, the keys still type. Selecting an Image Trace object no longer keeps widening the right-hand panel.

      Vector and PSD editors: shortcuts go to the editor, not the browser

      Ctrl (⌘ on Mac) + and − inside the editor no longer zoom the whole browser page and cut off the panels. Once you click into the editor, Ctrl and Alt combinations and F1 to F10 reach the editor instead of triggering browser actions such as save, print or find, and Ctrl+wheel zooms the canvas. Copy and paste work as before.

      Vector Editor: white type and fading gradients in Illustrator files

      White type in Illustrator (.ai) files whose text colours are stored as RGB no longer shows as dark grey. Gradients that fade to transparent now change colour as they fade, so patterns covered by a blend mode match what Illustrator and PDF viewers show.

      Compress HWP files: pictures no longer shrink inside their frame

      When an HWP file had large photos scaled down, the photos could appear small in the corner of their frame. The crop area and original size stored with each picture are now updated to match the new resolution, so pictures keep the same size and position as the original. Cropped pictures keep their crop, and HWPX files get the same fix.

    • Font finder: more accurate matches and a simpler screen

      Drop in a photo and the font finder picks a line of text and shows the result right away. Tap another line to search with it instead. It uses a new recognition model and better text reading, and it can match outline-only lettering. It now reads Cyrillic, Arabic, Persian, Thai, Devanagari and Greek, and adds Thai, Hindi and more free Chinese fonts.

    • Select and delete many PDF pages at once

      In the PDF editor page list, Shift-click selects every page in between and Ctrl-click (⌘ on Mac) adds or removes one page. Delete or rotate the selected pages in one go. The Delete pages and Rotate tools also take Shift-click ranges or page numbers such as "10-15, 20". Thanks for the suggestion.

    • JTD Editor: damaged or crafted files no longer use up memory

      Thanks to a private report from HIDANO Masaki (KUKAN project), we fixed a memory problem in ezpzfile-jtd. Length and count fields inside a file are now checked against what is really there, and a file that does not add up is refused with a notice. Files are checked in a separate worker before they open, so a bad file cannot freeze the tab. Very long documents show their first 2,000 pages. Engine: ezpzfile-jtd 0.3.2.

      PDF editor: editing Japanese text no longer leaves blanks

      When you type characters the original font does not have, the substitute font now matches the script: Japanese kana and kanji get a Japanese font, Hindi gets a Devanagari font. If even the substitute lacks a character, the editor tells you which one instead of saving a blank.

      The QR code generator now opens straight into the editor

      No more steps and page hops: the page opens right in the editor. Pick templates inside it, double-click any text to edit it in place, and use the bar above the selection for size, bold, color and alignment. Editing text tied to the QR, like the Wi-Fi name, updates the QR too.

      The mail viewer now opens Outlook MSG, PST, OST plus MBOX and EMLX

      Open .msg, .oft, .pst and .ost without Outlook, plus .mbox from Gmail Takeout or Thunderbird and .emlx from Apple Mail. Large PST files show their folders first and read mail only when you pick it, and the 1,000 mail limit is gone. PDFs now save directly without the print dialog, merged into one file or one PDF per mail. Also new: search by sender, date and attachment name, conversation view, .eml and CSV export, and an encoding menu.

      Fixed: "Open in editor" opening an empty editor

      In some setups (for example when the new window opened in the installed app), the editor opened without your file. The file now always arrives, and it still never leaves your browser.

      • New: PDF to CAD (DXF)

        Pull the lines, dashed lines and text of a drawing PDF into a DXF you can edit in AutoCAD. Pick the scale printed on the drawing for real sizes; you get one file per page and can open it in the CAD viewer right away. Direct DWG output is coming.

      • The font finder now finds and reads each line of text in a photo

        Even in busy photos like shop signs and station boards, the font finder now outlines every line of text and lists what it read as buttons. Tap one to search with that line; the largest line is picked first. Tilted lines and vertical Japanese are straightened before reading, and white text on colored backgrounds works too. The language is decided from what each line actually says, so Japanese photos are no longer taken for Korean. When you choose Vietnamese, Russian or Greek, fonts from independent makers now show their sample in that language.

        About 200 more free fonts from outside Google Fonts

        We added 202 free-for-commercial-use families from Velvetyne, Collletttivo, TeX Gyre and Latin Modern, Fontsource, Font Squirrel and others to the font list and image search, after reading each license. 44 of them cover Cyrillic, 40 Greek and 40 Vietnamese. Each detail page links to the license text, and downloads go to the maker’s official page.

        The font finder now reads Arabic, Persian, Vietnamese, Cyrillic and Greek

        We retrained the model that recognizes fonts with Arabic, Persian, Cyrillic, Greek and Vietnamese text. Connected Arabic script is matched by whole lines instead of single letters, and 61 free-for-commercial-use Arabic and Persian families are in, from Google Fonts plus Persian fonts such as Sahel, Shabnam and Samim. Free fonts with Cyrillic, Greek and Vietnamese letters can be filtered by language and found from an image too. Accuracy for Korean, Japanese and English is unchanged.

        New: compress HWP files

        Shrink HWP and HWPX files made heavy by photos, in the same format. Pick Print, Screen or Maximum, or press a target such as "Under 10 MB". The result is reopened by the document engine to show the first page, and the HWP editor save menu now has "Make file smaller". BMP and PNG screenshots, common in official documents, become JPG in both HWP and HWPX.

        The CAD viewer prints drawings to fit the paper

        Saving a large drawing such as a DWG as PDF no longer gives a blank page. It is now scaled to fit A4, A3, A2 or A1, and when model space holds many drawings, you get one page per drawing frame. For both PDF and PNG you can save only what you see or an area you drag over, and there is a print option for black lines and text.

      • The font finder now reads tilted and slanted text

        Text lines that are tilted because the photo was taken at an angle are straightened automatically, and letters that lean like italics are compared both straightened and as they are. If you photographed a sign or a page from the side, use Straighten angled photo, place the four corners, and the finder flattens it before matching. On tilted test images it now names the right font first about as often as on straight ones.

        The DOC, sheet and PPT editors no longer break on phones

        On narrow screens the menu names at the top no longer fold one letter per line. The menu wraps to two rows so every tab shows, sheets fit the screen width, and slides show larger.

        Open the editors and the most used tools straight from the home page

        The top of the home page is new. Under the tool search, icons open the PDF, HWP, DOC, sheet, PPT, PSD and vector editors, the CAD viewer, the font finder, file sending and the most used PDF tools in one click.

        Pick from 1,519 Google Fonts and the fonts on your computer in the editors

        The DOCX, XLSX and PPTX editors, the PDF Editor, the PSD Editor and the Vector Editor now list web fonts (1,519 Google Fonts). Only the names are listed; a font is downloaded from Google's server the moment you pick it and kept in this browser. Files that use one of these fonts download it on open and show as designed. In Chrome and Edge you can also load the fonts installed on this computer, paid ones included, after you allow it. Font files never pass through our server.

        The spreadsheet editor now saves formula results, and Korean text spaces evenly in slides

        Files saved from the XLSX Spreadsheet Editor now carry each formula's result next to the formula, so totals show in apps that do not calculate, such as mail previews and phone viewers. Formula cells no longer go blank for a moment after a save, and new cells use the workbook's own default font. In the PPTX Presentation Editor, text in a Korean font this computer does not have no longer draws with wide gaps between words. In Resize images, the new size in the file list now follows the settings as you change them.

        The office editors now open old .doc, .xls and .ppt files and AutoSave into the file you opened

        The DOCX, XLSX and PPTX editors now also open older formats (.doc, .xls, .ppt) and ODF files (.odt, .ods, .odp). They open as a converted copy, and saving makes a new file so the original stays as it was. Charts in documents, slides and spreadsheets open as real charts you can edit, and objects that cannot be read show as pictures. In Chrome and Edge, turn on AutoSave to save straight into the file you opened, and if you close without saving you can pick up where you left off next time. Big spreadsheets no longer freeze the page while they open, and the editors no longer get stuck while loading.

        Three new editors for Word, Excel and PowerPoint files, nothing to install

        The DOCX Document Editor, XLSX Spreadsheet Editor and PPTX Presentation Editor are live. Without an office suite you can open a file right in this browser, fix text, tables, formulas or slides, and download it in the same format. The file never leaves this browser, and a PDF you turned into Word or Excel opens in the editor with one click.

      • XDW editor: pages from a PDF keep their text, and the paper size shows correctly

        Pages brought in from a PDF used to become pictures only, so you could not search or select their text, and a PDF saved again had no text. The text now goes in with each page, under the picture, so it can be searched and selected after saving as XDW, and it stays in a PDF you save. Right after adding PDF pages, the properties panel kept showing the blank A4 page a new document starts from; it now shows the PDF's real paper size.

        Tapping a text box on a phone no longer zooms the page

        On iPhone and other phones, tapping the search box, a password field or the feedback box used to zoom the page in by itself. The page now stays the same size. You can still pinch to zoom. The Guide tab on the font finder also shows Korean, Japanese and Chinese labels upright on phones.

        • Ratings and feedback: tell us what got in the way

          When you finish a task, five stars appear in a corner. You can pick what fell short or leave a line, and each tool asks at most once a week. If a file will not open, Tell us about this file next to the error sends its type and version, and Send feedback in the left menu is there any time. Your file and its name are never sent.

          Built-in fonts like Malgun Gothic and MS PGothic are now matched exactly

          The font finder now compares letter shapes for the fonts built into Mac and Windows too. Built-in fonts such as Malgun Gothic, MS PGothic, Gulim and Meiryo, which used to lose out to similar free gothic fonts, now come first when the text is set in them. We also added about 70 paid Korean fonts sold on Noonnu Market, SEBANG GOTHIC 2.0 (free for commercial use) and HY Ulleungdo from Hanyang I&C.

          Font finder separates fonts whose commercial terms are unconfirmed, and a fix for one-line text

          Fonts that are free to download but whose commercial terms we have not confirmed now sit under their own price option, so Free for commercial use shows only fonts that carry that tag. The text step now says 1 line instead of 1 lines. In the CAD viewer, the note explaining how STL, OBJ and 3DS units were guessed is now shown in Korean and Japanese too.

          Extract now opens ALZip .egg and .alz files

          Open ALZip .egg and .alz archives in your browser without installing ALZip. Its maximum compression (AZO), LZMA, solid archives and AES or LEA passwords all work. For a split ALZip archive, choose all the parts together and they are joined; if one is missing, you are told which.

        • Chat Backup Viewer: read your exported chats like a phone screen

          Open chat exports from WhatsApp, Telegram, LINE, KakaoTalk, Slack, Facebook Messenger, Instagram DMs, iPhone Messages and Android SMS in your browser. Chats show up in bubbles with a feel close to each app, and even hundreds of thousands of messages scroll smoothly. Search, jump to a date, browse photos and files, see stats by person and hour, and save any period as HTML or PDF. It is read only, so chats can't be edited or written, and they never leave your device.

          A wider font list with filters, now with 9,700 free and paid fonts

          The font page now has a filter sidebar on the left and a wide font list on the right. Browse about 9,700 fonts, free for commercial use, paid from Sandoll Cloud, FONCO (Yoon Design), RixFont Cloud, AG FONT, Morisawa and Fontworks, and built into Mac and Windows, by price, language, style, weights and maker. We also retrained the model that recognizes fonts, now including the new Japanese fonts and the fonts built into Mac. It matches the right font more often on photo-like test images, and paid fonts now also have light and bold samples, so very thin or heavy text is recognized too.

          The font finder now knows over 1,400 Japanese fonts

          Free Japanese fonts shared directly by their makers grew from about 160 to about 270, and more than 1,000 Japanese fonts sold by Morisawa and Fontworks are now recognized from their public previews, with a link to the purchase page. Many Japanese fonts share their kanji and differ only in kana, so the finder now looks at kana shapes too.

        • CAD Viewer: SXF drawings and PCB fabrication data open

          SXF files from Japanese electronic delivery (.sfc, .p21, .sfz, .p2z) open in 2D with their layers, Japanese text and partial drawings, and measuring shows the real length at the drawing scale. PCB fabrication data (Gerber with drill files, ODB++, IPC-2581) opens with its layers in board order, a top and bottom view, a 3D board and a compare mode that shows what changed between two versions.

          CAD Viewer: old DWG files, IGES and Inventor open

          DWG files from AutoCAD 1.4 to R14 now open too, so every version from 1.4 to 2018 and later works. IGES models (.igs, .iges) open in 3D with their surfaces and solids, and Inventor parts and assemblies (.ipt, .iam) open with each part in place: drop the whole assembly folder or pick the files together.

          Paid fonts and Mac & iPhone system fonts now appear in results

          Paid fonts now rank alongside free ones with a preview, and can come up as the best match. They link to the seller with "View at seller", where you can try your own text. The finder also recognizes 80-plus fonts built into Mac and iPhone, such as SF Pro, Helvetica, Avenir, Apple SD Gothic Neo and Hiragino.

          Font finder now points to paid fonts too

          Using the public previews of more than 4,000 paid fonts on Sandoll Cloud, the finder now tells you when the font in your image looks closer to a paid font than to anything in our free list. Similar paid fonts are listed with their maker and link straight to the seller's purchase page. On our test images the right paid font comes first nearly eight times in ten and is in the top five more than nine times in ten.

        • CAD Viewer: iCAD SX files open

          The CAD Viewer now opens iCAD SX .icd files without iCAD: 3D solids in 3D, and 2D views such as front and plan views as drawings with their dimensions and text. Files from V5L3 and V7L5 were tested.

          Free fonts now link straight to their official sources

          We rebuilt the list of over 900 Korean and 160 Japanese fonts around government-approved fonts and each maker's official download page. Every font shows its maker, source and license, and downloads come straight from the government or maker server without leaving the page. Image search now adds a recognition model we trained on over 5,000 fonts: on our test images the first result is right 74% of the time, up from 62%, and it also works on handwriting whose letters are hard to separate.

          CAD Viewer: STEP and Jw_cad JWW files open, and the viewer is now in Japanese

          The CAD Viewer now opens STEP parts and assemblies (AP203, AP214, AP242, also .stpz) with smooth curved surfaces, their edges and a tree of the assembly to switch parts on and off. Jw_cad drawings (.jww) open in 2D with their layer groups, Japanese text and line types, and measuring shows the real length at the layer group scale. The viewer and its page are also available in Japanese.

          Every font now has its own page

          Open a font from the list to see a large preview, try your own text, compare weights and sizes, check every letter, read the license and browse similar fonts on one page. We also removed a few noonnu fonts that showed empty boxes and fixed previews that mixed up shapes between noonnu fonts.

          Font finder: drop an image to see which font it uses

          Browse more than 2,000 free fonts for commercial use from Google Fonts and noonnu, in Korean, Japanese, Chinese and Latin. Drop a screenshot or photo anywhere on the page and the closest fonts appear, matched letter by letter. The image never leaves your browser.

          The RVT viewer is now the CAD Viewer: one address for drawings and models

          The CAD Viewer opens DWG, DXF and DWFx drawings with their layers and layouts, RVT and RFA files and IFC models in 3D with their properties, SketchUp (SKP) models, and FBX, OBJ, STL, glTF and 3DS models, all at one address. The format is recognised from the file itself, files are still read only in your browser, and Try a sample lets you pick a file in each format. The old RVT viewer address now leads here.

        • RVT viewer: try a sample, roof textures, faster detail switch

          No .rvt file at hand? Press "Try a sample" on the start screen to open a school building model (by Ali Kücükavci, CC BY 4.0). Textures and surface patterns now keep their real-world size on slanted faces such as roofs instead of stretching. Switching the detail level (coarse, medium, fine) no longer reads the file again, so it is several times faster, and your camera and section box stay where they were.

          RVT viewer: realistic mode and two-letter shortcuts

          Turn on Realistic to see brick, concrete, wood, grass and other materials with textures at real-world size. The file does not contain texture images, so look-alike CC0 textures are chosen from the material names. Category names are now readable, and you can pick the detail level (coarse, medium, fine) and show surface patterns. Two-letter shortcuts such as HH, HI, ZA, VV and SB work too.

        • New: RVT viewer

          Open .rvt building model files in 3D in your browser with nothing to install. Orbit and zoom, show or hide levels and categories, measure distances, angles and areas, and cut the model with section planes. Open sheets and save them as PDF, and export walls, doors, windows and rooms to Excel. Tested with files from versions 2022 to 2026, and the file never leaves your device.

        • Korean file names no longer show as boxes in the PSD and vector editors

          Files with Korean names opened on a Mac showed boxes in the tab and title bar, because macOS passes those names with the letters split into their parts. The editors now put them back together first. On a page in another language, the editors also fetch the font for your browser language (Korean, Japanese or Chinese) so that text shows.

        • Edit Thai and Vietnamese text in PDFs, and keep folders when extracting

          Double-clicking Thai or Vietnamese text in a PDF and retyping it used to save with tone marks or letters missing. That is fixed: letters the original font lacks are filled in with a Thai or Vietnamese font. In the archive extractor, "Download all (ZIP)" now keeps the original folder structure.

          Open many EML files at once as a mailbox

          Drop several .eml files, a folder or a ZIP and the viewer turns into a mailbox, with the list on the left and the selected email on the right. Search by sender, subject or body, filter by folder, save every listed email into one PDF or download all attachments as a ZIP. Up to 1,000 emails at a time, and the files never leave your device.

        • Background removal gets an HD mode and a brush

          Each photo is cut out as soon as you add it, with a slider to compare against the original. If hair or fur looks rough, redo it in HD and touch up the rest with the erase and restore brushes. Swap the background for a colour or your own photo, then save as PNG, WebP or JPG at full size, or everything as one ZIP.

          Editors now show their loading progress

          The PDF, HWP, JTD, XDW, PSD and vector editors show a progress bar with the megabytes received and the current step while they load. On later visits they say they are loading the saved editor, and if the network stalls you get a note and a retry button instead of a frozen screen.

        • Compress a photo or PDF to an exact size in KB

          Type a limit such as 20-50 KB or under 200 KB. Photos come out at the highest quality that fits, and PDFs are shrunk as gently as possible with the text kept selectable where it can be. Presets cover exam photos (200 x 230 px) and signatures (140 x 60 px), and minimum sizes are respected. It works on phones and files stay on your device.

          Vector Editor now saves as AI by default

          Save As now picks PDF-compatible AI instead of the editor's own .vectorcraft format. A new document is suggested as "Untitled-1.ai" and an opened .ai file keeps its name, so the saved file opens straight in Illustrator.

        • New Vector editor: open AI files without Illustrator

          Open an Illustrator .ai file in your browser with its layers, groups and paths intact, edit it, then save it back as AI, SVG or PDF or export PNG and JPG. SVG, EPS and PDF files open too. The menus, shortcuts and panel layout follow Illustrator. It is built on the open-source VectorCraft with a light theme and a Korean interface added, and like our other editors it never sends your file anywhere. With this many panels it runs on computers and tablets, not phones.

        • New PSD editor: open PSD files without Photoshop

          Open a PSD in your browser with its layers, masks and adjustment layers intact, edit it, then save it back as PSD or export PNG and JPG. The menus, shortcuts and panel layout follow Photoshop. It is built on the open-source PhotoCraft with a light theme added, and like our other editors it never sends your file anywhere. With this many panels it runs on computers and tablets, not phones.

          You can now install EZPZ File as an app

          Install it straight from your browser, with no installer to download. It works on Windows, Mac, Android and iPhone, and once installed it opens in its own window from your desktop or Dock. On computers where you install it with Chrome or Edge, a double click on a PDF, HWP, Ichitaro or DocuWorks file opens it in the editor. Save the editors once and they open even without internet, and on Android you can share a file from your email or chat app straight to EZPZ File.

          Vietnamese and Thai text now saves correctly from the PDF editor

          Vietnamese letters with tone marks and Thai text looked fine in the PDF editor but came out as empty boxes in the saved PDF, because the font we embedded did not have those characters. The editor now embeds Be Vietnam Pro for Vietnamese and Sarabun for Thai, so the text stays as you typed it. In the EML viewer, an email attached to an email (a forwarded original, or the postacert.eml inside an Italian PEC) opens with "Open this email", and dates follow the page language.

          The EML viewer now saves emails as PDF

          Open an EML file and press "Save as PDF". The print window shows the subject, sender, recipients, date and attachment list above the message, and you pick "Save as PDF" as the destination. Images inside the message go in, and so do image attachments such as photos and scans, placed after the body. The text stays text, so you can search and copy it in the PDF. Images loaded from the web can tell the sender you opened the email, so the viewer asks before including them.

          Scanned PDFs keep their text in the PDF tools

          Office scanners often save a page as a light background photo with the text and lines in a separate black-and-white layer. The PDF editor and PDF to image could not decode that layer, so only the faint background showed, and compressing such a PDF by turning its pages into images dropped the text as well. The text and lines now come out as in the original.

        • QR codes for contacts, chat apps and payments

          You can now make a contact card (vCard) that saves straight into the phone, an email or text with the recipient filled in, a map location, a calendar event, and chat codes for WhatsApp, LINE and Telegram. Payment codes follow each country's published standard for Pix in Brazil, UPI in India, PromptPay in Thailand and SEPA transfers in Europe (EPC QR, also called GiroCode), and a new payment sign template prints the payee and amount under the code. The money goes from the payer's bank straight to yours and never passes through this site.

          The QR code generator now makes Wi-Fi codes, logo codes and printable signs

          Pick what the code is for (Wi-Fi, a menu, reviews, social, a business card and more) and you get sign templates that fit, then finish them in a full-screen editor with text, shapes, images and any Google Font. Enter a network name and password for a code guests scan and join, add a logo in the middle or change the dot and corner shapes, and every change is checked to make sure it still scans. Downloads come out at the exact size you pick as PNG, print-ready PDF or SVG, and codes with Korean, Japanese or emoji no longer come out broken.

          New documents in the JTD editor save as Ichitaro files

          A document you start from scratch in the Ichitaro (.jtd) editor now saves as .jtd. It gets the page a new Ichitaro document has (A4, 40 characters by 40 lines), and Ichitaro Viewer opens the saved file. Bold, size, colour and underline added to text that had no formatting used to show as plain text in Ichitaro Viewer; they show now, and italics are saved as well.

          Start a new document in the PDF and XDW editors

          The PDF editor has a New document button on its start screen: begin with a blank page, add text, pictures or a signature, and save it as a PDF. The DocuWorks (.xdw) editor starts from a blank A4 page as well, and PDFs or photos dropped on it with nothing open become a DocuWorks file of their pages. Saving an XDW now writes the whole file anew, so deleted pages and annotations no longer stay inside it, and Japanese or Russian text added in the PDF editor now shows in the saved PDF.

          A how-to video in the JTD editor

          The Ichitaro (.jtd) editor now has a four-minute how-to video under the drop area, narrated in Japanese, on both its Japanese and English pages.

          A how-to video in the XDW editor, and JTD and XDW guides in English

          The DocuWorks (.xdw, .xbd) editor now has a four-minute how-to video under the drop area, narrated in Japanese, on both its Japanese and English pages. The guides to opening Ichitaro (.jtd) and DocuWorks files are now in English too, and both editors link to them.

        • Make and open .ezpz archives in your browser

          A new tool packs files and folders into one small .ezpz and opens the ones you receive. Once it is packed, the result page lists everything inside, folders and all. A password hides even the file names, and whoever you send one to can open it on the same page with nothing to install. The engine is open source under the MIT License, and the format is fixed at version 1, so files made today will keep opening in later versions.

        • Ichitaro and DocuWorks editors, in Japanese and English

          Ichitaro (.jtd) and DocuWorks (.xdw, .xbd) files, common in Japanese offices and government paperwork, now open in the browser for editing, with nothing to install. Ichitaro documents save back to .jtd or download as Word or PDF, and in DocuWorks documents you can add sticky notes, highlights and stamps or rearrange pages before saving as DocuWorks or PDF. Both editors run in Japanese and English, and the old JTD viewer address now leads to the new editor.

          MCP server 0.4 turns PDFs back into Word, Excel or HWP and reads Ichitaro and DocuWorks

          The EZPZ File MCP server that AI agents call can now rebuild a PDF as a Word, Excel, HWP or HWPX file, and change words inside a PDF while keeping the original font. It reads Ichitaro and DocuWorks documents and converts them to PDF, opens iPhone HEIC photos, and lists what a file records about itself, such as where a photo was taken or who wrote a document. When an HWP file will not open, the server says why, and an engine that stops on one file no longer fails the files after it.

        • HWP conversions now start in every Edge setup, and files that will not open tell you why

          In some Edge setups the HWP and HWPX conversion engine failed to start. That is fixed. When a Hangul file will not open, the message now says why, because the next step differs: a file locked by company or government document security (DRM), a Word file or PDF that only has an .hwp name, or a download that was cut off. If the engine stopped partway through a document, later conversions in the same tab used to fail too; that is fixed, and the HWP editor shows the same guidance.

        • Pictures, text boxes, connectors and equations go into HWP documents in one step

          A picture now lands at the cursor as soon as you pick the file, and a new text box is ready for typing the moment you draw it. Choosing a shape or connector shows what to do next right on the page, connectors can be drawn anywhere, and an end dropped on a dot at a shape's edge stays attached when the shape moves. Equations can be resized by dragging a corner and moved by dragging them to another spot in the text, and the text wrapping icons in Object Properties are now easy to tell apart.

          Fixed reordering long PDFs, locked files in Merge, and three smaller issues

          In PDFs longer than 120 pages, dragging a page used to drop every page after 120 from the saved file. That is fixed. Merge now stops a password-protected PDF and asks you to remove the password first, instead of merging its pages as blank sheets, and Protect asks you to unlock an already protected file first. The .docx text tool now explains old .doc and password-protected files properly, and HWPX to text no longer repeats table text.

          Tool pages now link to related articles, plus 9 new articles

          Every tool page now ends with the articles that cover that tool. We wrote new ones for the tools that had none: merging PDFs, deleting and reordering pages, removing a password you know, PDF to image, iPhone HEIC photos, shrinking a photo for an upload form and opening a .docx without Word, each in 20 languages. In English and Korean we also added a ranking of free HWP editors and a guide to turning HWPX into HWP when a Korean office asks for it.

          You can now share tools and articles

          Tool cards, tool pages, the result screen, the PDF and HWP editors and every article now have a share button. On a phone it opens your phone's share sheet, so you can send the link with any messaging app. On a computer you can copy the link or post it to X, Facebook, Reddit or LinkedIn. Only the page address is shared, never the file you worked on.

          Articles now start from the problems people actually hit

          We rewrote 22 articles around the error messages and situations people describe in Q&A boards and comments. A zip that says it is empty, a password zip that fails with 0x80004005, an .hwpx file that suddenly shows up in your inbox: each article now answers the exact case people search for. The changes cover 10 languages, including English, Korean, Japanese and German.

          Password-protected archives now ask for the password

          A password-protected ZIP used to come out as broken files, with no question asked and no error shown. The extract tool now asks for the password, extracts right in your browser and tells you when the password is wrong. Password-protected 7z and RAR archives work the same way.

        • HWP editor is live

          You can now open HWP and HWPX documents in the browser and edit them without Hancom Office. Fix text, tables and formatting, then save as HWP or HWPX, or print to PDF. The document never leaves your computer.

          JTD viewer for Ichitaro documents

          Documents made with Ichitaro, the Japanese word processor, now open in the browser without installing anything. Body text and tables, merged cells included, appear in their original order, and you can save the result as Word, text or PDF. The viewer is available in English and Japanese.

          Preview your converted file before downloading

          When a conversion finishes, the result page now shows the file itself. Page through PDFs and HWP documents, zoom into images or compare them with the original, and read spreadsheets as a table. You no longer have to download a file just to check that it came out right.

          Excel dates stay dates in CSV and JSON

          Date cells used to come out as serial numbers such as 46024 when an Excel file was converted to CSV or JSON. They now come out as dates like 2026-01-02, and a custom date format set in Excel keeps its look. Saving back to Excel still opens them as dates.

        • PDF to Word and Excel now speak all 20 languages

          The PDF to Word and PDF to Excel screens used to show only Korean and English. Buttons and messages now appear in the other 18 languages as well, from Japanese to Spanish and German. Wherever you arrive from, the tool works the same way in your language.

        • PDF unlock now tells you when the password is wrong

          A mistyped password used to produce a misleading message saying the file was damaged or in the wrong format. The tool now says plainly that the password does not match, and it also tells you when a file has no password at all. You can simply try the password again instead of doubting the file.

        • More reliable saving in the PDF editor

          We fixed three cases where an edited document failed to save or showed a blank page in some PDF viewers, such as text set in a substituted font or an image whose file extension did not match its real format. If saving fails, the editor now tells you why, and it asks before you close the tab with unsaved changes. Your edits are much less likely to get lost.

          Edit Chrome "Save as PDF" documents in place

          PDFs made with Chrome's "Save as PDF" split each line across several fonts, so edited text used to be stamped on top of the original. The editor now recognises those pieces as one line, removes the original text and writes the new text in the same spot. This matters most when you save a web page as PDF and then need to fix it.

        • A new layout with a side menu, tool search and pick by format

          A side menu now takes you to all tools, the editors and the articles from any page. Type what you want to do and the search finds the right tool, and on the home page you can start from your file's format to see only the tools that fit it.

        • PDF editor is live

          Double-click existing text in a PDF to edit it in place, or double-click an empty spot to add new text. Add signatures, images, highlights and shapes, add or remove pages, then save a new file. It works in all 20 languages and nothing is uploaded.

          The PDF editor works on phones

          Double-tap existing text to edit it, and the keyboard now appears on iPhone too. Pinching zooms the document only, not the toolbars, and you can drag signatures and text boxes with a finger.

          PDF tools no longer fail on some browsers

          On about 18% of browsers, including Samsung Internet and slightly older versions of Chrome and Safari, turning a PDF into images or previewing its pages ended in an error. We switched to a build of the PDF engine that runs on more browsers, so those features work again. You do not need to change browsers.

          When a conversion fails, you now see why

          Almost every failure used to end with one sentence asking you to check the file. Tools now tell apart eleven causes, such as a password, a scan with no text or running out of memory, and suggest what to do next. This applies to every tool on the site.

          EZPZ Drop is now in the menu

          EZPZ Drop, for moving files too big for email from your phone to your computer, now opens straight from the site. Files are encrypted before they leave your device and the key lives only in the link, so we cannot open them. When both devices are open, files go directly between them without passing through our server.

        • Convert HEIC photos to JPG or PNG

          iPhone photos in HEIC often will not open on Windows or in Chrome. You can now turn them into JPG or PNG that open anywhere, and photos taken sideways come out the right way up. Your photos are not uploaded.

        • PDF compression now shrinks only the photos by default

          Most of a PDF's size is usually its photos. The default mode now re-encodes just the photos at a sensible resolution and leaves the text alone, so you can still select and search it afterwards. Turning whole pages into images is still there for when you need the file even smaller.

        • Turn PDFs back into Word, Excel, HWPX and HWP

          Four new tools read the text and tables in a PDF and rebuild them as an editable document. Tables come across as tables, and stamps and images on the page come along too. Use them when only the PDF is left and you need to edit it again.

        • See what is hidden inside a photo or document

          Drop a photo to see when it was taken, the camera and lens, and the location written into the file on a map. Drop a document to see its author, how many times it was saved and the total editing time. Check what travels with a file before you send it.

        • Extract 7z and RAR archives

          The extract tool used to open only ZIP files. It now handles 7z, RAR, tar, gz, xz and bz2 as well, and file names in Korean or other scripts come out intact. There is nothing to install.

          A guide to using EZPZ File from AI agents, in 20 languages

          One page now explains how to add EZPZ File as a tool to AI agents such as Claude and Cursor. After a single install command, an agent can read and convert documents, edit PDFs, edit images including background removal, and make QR codes. The guide is available in all 20 languages.

        • First release of the MCP server for AI agents

          AI agents used to write, debug and check conversion code every time they needed to change a file, and the MCP server turns that into a single call. The first version offers four tools: reading documents, converting documents, editing PDFs and inspecting PDFs. Like the site, it does the work on your own machine.

          HWPX to HWP, and HWP to HWPX now keeps tables and images

          A new tool turns HWPX back into HWP for older Hangul versions and systems that only accept .hwp files. The HWP to HWPX direction now runs on the same engine, so tables, images and equations that used to be dropped come across as well.

        • Convert HWP and HWPX documents to PDF

          HWP and HWPX documents can now be turned into PDF without Hancom Office. Pages are written as real text rather than pictures, so you can select and search the text in the resulting PDF. Only the fonts the document actually uses are downloaded, which keeps it quick.

        • Remove photo backgrounds

          Drop a photo and the background is removed, leaving the person or object on a transparent PNG. Even the model that finds the background is downloaded from our own site and runs in your browser, so the photo is never uploaded anywhere.

          Strip hidden data from photos and PDFs

          Remove the location and device details stored in photos and the author record stored in PDFs. Only the metadata is cut out and the photo is not re-compressed, so the quality stays exactly the same. It is worth doing before you post a photo or send a document.

        • Make QR codes

          Turn links, Wi-Fi details or plain text into a QR code and download it as PNG or SVG. You can change the colours, and what you type never leaves the browser.

          Articles on everyday file problems, in 20 languages

          We started publishing articles that solve common file problems, such as a PDF too big for email or scanned pages that need to become one file. The article section opened in all 20 languages on the same day.

        • EZPZ File launches

          We launched with 20 tools for merging, splitting and compressing PDFs, converting images, and working with Hangul and Excel files. Everything runs inside your browser, so files are never uploaded, and there is no sign-up or usage limit. The site opened in 20 languages from day one.