ORBITThe research pipeline, the templates behind it, pricing, and what happens to your manuscript.
ORBIT (Optimized Research Blueprint and Instant Templating) is a research publishing pipeline — Discover, Check, Submit, Present — ten tools that share one credit pool. Literature and Dataset Discovery find what to build on; Readiness Check tests a manuscript against a journal's stated rules; Format Paper, Journal Suggestion, Submission Letters and Conference Deadlines get it submitted; Paper to Slides turns it into a talk. Four tools are free, six cost 1–3 credits.
You need to sign in with your Google account to use the tools. Sign-in is what lets us track your free monthly credits and any paid credits you've purchased. Browsing the site, reading templates, and exploring pricing does not require an account.
Yes. Every account starts on the free Base plan: 5 credits per calendar month at no cost, and no card is required. They reset on the first of every month. Four tools take no credit: Literature Discovery search, Dataset Discovery and Journal Suggestion (10 searches a month each on Base) and Conference Deadlines (no limit). The rest cost 1–3 credits: 1 for a standard format, a readiness check against a style system, a cover letter or a deck outline; 2 for a named-journal readiness check or a custom-template format; 3 for a narrated deck. Your free credits cover every one of them. The exception is a university email address: until its institution sets up ORBIT, it joins a waitlist for that institution instead of getting free credits, and it can still buy credits.
Yes. ORBIT works on mobile, tablet and desktop browsers — you can upload a .docx, pick a template, and download the formatted paper from a phone. The readiness check also runs on mobile, though it analyses the document on your own device, so a long manuscript will take longer on a phone than on a laptop.
So far, researchers from 12 countries have signed up: India, United Kingdom, Nigeria, South Korea, Pakistan, Australia, Iraq, United States, Bangladesh, South Africa, Lebanon and Jamaica. Each country is backed by at least one real account. ORBIT publishes no user counts and names no one without their permission.
Yes. YourStory named ORBIT in its VibeCode30 2026 Top 30, picked from 207 applications to the first edition of its list of live products built mostly with AI tools in India. The winners were announced at TechSparks 2026 in Bengaluru on 14 October 2026. ORBIT's profile is on pages 28 and 29 of YourStory's report, which is linked in the footer of every page.
ORBIT currently accepts .docx files (Microsoft Word format) as input and produces a formatted .docx file as output. PDF input and LaTeX export are not offered.
Most conversions finish in seconds. Processing time varies with document length and complexity — a short paper is close to instant, a long one with many figures takes longer.
After the conversion is complete, a download button appears in your browser. Click it to download the formatted .docx file directly to your device. There is no email delivery — the download is immediate.
Never. ORBIT only changes the formatting — fonts, margins, column layout, heading styles, citation format — not a single word of your research content is altered.
It is transformed in memory and sent straight back to you — no copy is written to disk or kept anywhere. What we record is the metadata around the job: the template, the file name and size, and the timestamp.
ORBIT preserves all text, tables, figures, and inline equations from your original .docx file. For highly complex mathematical notation (LaTeX-style), we recommend a quick review after conversion to ensure correct rendering.
ORBIT has 21 templates. Publishers: IEEE, ACM, Elsevier, Springer LNCS, Nature, ACS, MDPI, Wiley, Taylor & Francis, Frontiers, PLOS ONE, IOP Publishing, Cell Press, BMC and SAGE. Style guides: APA 7th, MLA 9th, Chicago, Harvard (Cite Them Right), AMA and Vancouver. Each one's numbers come from the publisher's own Word template or author page, and the Templates page lists every value with its source. Where a publisher sets no layout (Wiley, SAGE, Elsevier at submission), the page says the layout is ORBIT's.
Yes. Select 'Custom Template' and upload your journal's own .docx or .dotx template. ORBIT reads its page size, margins, columns, body font and size, title and heading sizes, line spacing, alignment, indent and line numbers, and shows you what it read and what it had to assume before you convert. It applies that layout; it does not copy logos, headers or text boxes from the template.
14 of the 21 were read from the publisher's own template or author page and checked against it: IEEE, ACM, Nature, ACS, MDPI, Taylor & Francis, Frontiers, PLOS ONE, IOP Publishing, Cell Press, BMC, APA, MLA and AMA. The other 7 (Elsevier, Springer, Wiley, SAGE, Chicago, Harvard and Vancouver) are honest about their footing: the publisher or style guide sets no layout at submission, or keeps it behind a bot block or a paid manual, so the layout is ORBIT's own and the template says so. We call nothing verified until it has been read against its source, because when we first did that for IEEE, ACM and Springer, all three had been wrong. Each template links to the document its rules came from. Always do a final visual review before you submit.
Because we could not read the publisher's own specification for them. A template that is not confirmed still formats your paper — it just means the numbers behind it have not been checked against a source we can point you at. Every card states which of the two is true on its face, and we would rather say it plainly than show a green tick we haven't earned. The link in the full specification takes you to the publisher's page so you can check it yourself.
Yes. We regularly add new templates based on user demand. If your journal format is not currently available, contact us via the Contact page or WhatsApp and we'll consider adding it.
It reads a finished manuscript against a journal's own submission guidelines and reports what does not comply yet — page and margin geometry, required sections and their order, title and keyword limits, citation style, figure and table caption rules, language checks, and metadata like author affiliations. Every failing check is paired with the specific change that fixes it, and you can download the whole report as a PDF.
No — and this is the part worth understanding. The .docx is unzipped and analysed by JavaScript running in your own browser, on your own processor. There is no upload step because there is nothing to upload to. Only your score, the journal name and a one-way fingerprint of the title are recorded against your account. The fingerprint cannot be turned back into the title; it only lets a re-check of the same paper within 24 hours go free. This is how the tool is built, not a policy promise.
It means the journal does not publish a rule for that check anywhere we can source, so it is excluded from your score rather than counted as a pass. Most compliance tools score every rule whether the journal states one or not, which quietly turns silence into a passing grade. A check here has three outcomes — pass, fail, and not stated — and only the first two count toward your score.
Yes. Paste the URL of that journal's author-guidelines page and ORBIT reads the rules straight off it. Anything the page does not state stays "not stated" in your report rather than being guessed at. A profile read this way is labelled as scraped and unverified by a human, so treat the report as a first pass.
Coverage is deliberately narrow and scoped to computer science, AI and machine learning. Eight journals were transcribed by hand from the publisher's own author-guidelines page. Three more (ACM Computing Surveys, ACM Transactions on Information Systems and Heliyon) are available but fall back to their citation style's defaults, because their publishers block automated reading of the guidelines page. They are labelled as unverified in the picker. We would rather cover a small number accurately than claim fifty and invent the numbers.
Usually. A readiness check against a built-in style system costs 1 credit, the same as a standard format — both from the same pool. A readiness check against a named journal costs 2, same as a custom-template format; a narrated deck costs 3. Your free credits cover every price.
A tool that makes the figures for a paper: charts from your tables and diagrams from your text. Upload a CSV, TSV or XLSX and it reads the columns (what is a category, what is a number, what is a date) and offers the figure types that actually fit that shape, 45 of them across comparison, distribution, relationship, trend, composition, matrix, model-evaluation, clinical and flow families. Or paste text and it draws a diagram. You pick one and export it.
Yes. Paste a methods paragraph, a list of themes or a project plan (15 to 1,500 words) and it draws one of 15 diagram types: a flowchart, sequence, cycle, mind map, timeline, tree, fishbone, pipeline, hypothesis model, comparison matrix, 2×2 matrix, Venn diagram or graphical abstract, or a PRISMA 2020 or CONSORT 2025 flow diagram drawn to the official template, or it picks the one that fits. You can then edit it in place: retype a box, add or delete boxes, draw arrows and drag them, for free. In PRISMA and CONSORT, a count your text does not give is shown as n = ? and never worked out for you, and counts that do not add up are flagged. The labels are phrases from your text, and no number, date or title appears unless your text states it. A diagram that would break this rule is not shown and not charged for. A tree can be switched to a mind map or a flowchart, and labels can be shortened twice, all for free. The text is sent to our AI to find its structure and is not stored.
1 credit per table, charged when the table is read, and 1 credit per text turned into a diagram. After that every figure type, every diagram layout and every export is free — try all 45, re-export at a different size, change your mind, none of it costs another credit. If no figure type fits the table, nothing is charged. Re-running the same table within a day returns the same analysis without charging again.
SVG, PNG at 300, 600 or 1200 DPI, and PDF. The PDF is real vector output — text stays text and lines stay lines, so it scales and prints without softening. JPEG, TIFF and EPS are not offered: JPEG is lossy and wrong for line art, and TIFF and EPS are legacy submission formats that most publishers now accept a PDF in place of. If a journal insists on TIFF, open the PDF and export one.
Computed, and refused when they cannot be. Error bars are drawn only from a column you identify as error or standard deviation — never inferred from a spare numeric column. Regression lines, confidence bands, box-plot quartiles and histogram bins are calculated from your values with published methods, and every statistical function is tested against a worked example from the literature rather than against its own output. Where the data does not support a figure type, the type is not offered. ORBIT draws your results; it does not run your statistical tests. A significance bracket shows the label you give it, forest plots pool by fixed effect only, and Kaplan-Meier curves carry no log-rank p-value. Those choices belong to you and your statistician.
The table is parsed in your browser. What crosses the network is the schema (column names, inferred types, row count) and never the cell values, which is enforced by a test rather than by a promise. Figures are rendered in the browser from that schema and your local data. Nothing is written to disk. For a diagram, the text you paste is sent to our AI to find its structure, and it is not stored.
The default palette is neutral and colourblind-safe, chosen for print rather than for a screen. Brand palettes exist as an option and are never the default — a figure in a paper belongs to the paper, not to the tool that drew it.
A tool that resolves every reference in your bibliography against Crossref, OpenAlex and DataCite (arXiv preprints included) and reports what it finds: DOIs no registry holds, entries that no longer resolve, entries whose author, year, journal or page numbers disagree with the published record, and papers that have been retracted. In manuscript mode it also reads the body text, so it can name citations that have no entry in your list and entries in your list that the manuscript never cites.
Two things. It reads the manuscript, not just the list — so it catches the in-text mismatches a list-only checker cannot see, which is where most real reference errors live. And it checks the list against your target journal's own referencing rules, not a generic style guide. It is not a hallucination detector; it resolves what you wrote against the published record and tells you where the two differ.
1 credit for a pasted list and 2 for a full manuscript audit, charged once per check. Restyling the corrected list into any of the twelve supported styles is free, and so is the .docx download — as many times as you like. If the check finds nothing at all, it is free: that is most often a vendor outage, which is the worst possible moment to charge. Re-running the same list on the same day returns the original receipt rather than charging twice.
We tested it against 30 papers known to be retracted, and it flagged all 30. It reads Crossref's update records, which is where publishers register a retraction. An expression of concern, where a journal warns about a paper without retracting it, is reported as its own finding and never counted as a retraction. A dataset DOI is looked up at DataCite, where datasets are registered, instead of being reported as not found.
It is flagged and left exactly as you wrote it. Nothing is invented, corrected on a guess, or quietly dropped. A reference ORBIT could not find in Crossref may still be perfectly real — a book, a thesis, a preprint, a conference paper that was never deposited — and the report says so rather than pretending to know better.
It flags the two ways they usually show up, but a flag is not a verdict. An invented reference normally has no record in Crossref, OpenAlex or DataCite, so it comes back as unresolved, and a DOI that does not exist is refused. An invented reference that borrows a real DOI shows up as a mismatch between what you wrote and the paper that DOI belongs to. Some real works are unresolved too, such as books, theses and papers that were never deposited, so treat a flag as a reference to check by hand, not as proof it is fake.
As accurate as Crossref, which is to say usually excellent and occasionally incomplete. Publishers deposit their own records and some deposit them badly — a missing page range, an initial where a full name belongs, a journal title abbreviated differently. Every field is therefore flagged individually with what the record says and what you wrote, so you can judge a disagreement rather than accept a correction.
Yes. Download it as a .bib file, or press Open in Overleaf and the checked bibliography opens there as a project with a minimal main.tex that prints it. The file is built in your browser; pressing Open in Overleaf is what sends it, and it sends the reference list only. Every reference is in the file, including any we could not verify — those carry a note saying so, so nothing goes missing from your bibliography without telling you.
A .docx manuscript or a pasted reference list, up to 300 references in one run. PDF is not supported: pulling a reference list out of a two-column PDF is materially harder than out of a .docx, and a half-read bibliography is worse than none. Export your manuscript as .docx and it reads cleanly. Over 300 references the check refuses rather than silently truncating, because a shortened audit that does not say it was shortened is the one failure mode worth refusing.
On the free Base plan you get 5 free credits and 10 searches in each free tool every month. Starter (₹799 a month or ₹8,630 a year) and Ultra (₹1,499 a month or ₹16,190 a year) give 10 and 25 credits every 30 days, 20 and 50 searches, and a Library; they renew on their own until you turn auto-renew off, or you can buy a 30-day pass that does not renew. Top-ups cost ₹150 a credit in packs of 5, 10 or 15 and do not expire while the account is active. Ultra also adds priority support, with a reply within 1 working day. Buying a top-up resets your free-tool searches for the cycle. The exception is a university email address: until its institution sets up ORBIT, it joins a waitlist for that institution instead of getting free credits, and it can still buy credits.
Top-up credits do not expire while your account is active, so you can use them months after you buy them. An account nobody has signed in to for 12 months is deleted together with its balance, but only after we email you 30 days in advance, and signing in once keeps everything. Plan credits are different: Base credits reset on the 1st, a paid plan's credits every 30 days (on a yearly plan too), and neither rolls over.
Paying in rupees, Starter and Ultra renew monthly or yearly (yearly is 10% off) by UPI Autopay or card through Razorpay. Turn auto-renew off at any time under Plan & credits in your dashboard: nothing more is charged, and you keep the plan until the end of what you paid for. The unused part is not refunded. If you would rather pay by hand, buy a 30-day pass instead; it does not renew, and buying it again adds 30 days, up to 90 days ahead. Outside India, plans are sold as 30-day passes through PayPal. When a plan ends you are on Base again, and your Library stays readable.
Open Plan & credits in your dashboard and pay through Razorpay's checkout. In Indian Rupees: UPI, card, net banking or wallet, whichever you prefer. From outside India: switch the price to US Dollars and pay with PayPal in the same window. A plan switches on, or credits land on your balance, the moment the payment is confirmed, and an invoice appears under Billing in your dashboard. ORBIT never sees your card number, UPI PIN or CVV; those go to Razorpay or PayPal, not to us. If Razorpay checkout is unavailable, a manual UPI fallback is shown instead — you pay the UPI ID displayed and send the confirmation on WhatsApp with your account email, and credits are added after a person verifies it, usually within a few hours. That fallback needs an Indian bank account.
All purchases are final. No refunds are provided once a payment is completed, because delivery is instant and digital. The exceptions are payment faults rather than changes of mind — a duplicate charge, or money taken with no plan or credits applied — and those are refunded through Razorpay to the original payment method. Use the free Base plan to try every tool before you buy anything. See the full Refund Policy for details.
Yes, and it is worth being precise about how. A readiness check never sends your manuscript anywhere: it is analysed in your own browser. A reference check sends only the references it reads from your file, to look them up. A formatting conversion or a slide deck sends the .docx over HTTPS to be processed in memory and sent straight back, and no copy is kept. No stage stores your document, no person reads it, and nothing you send is ever used to train an AI model.
Your Google or LinkedIn identity (email, name, profile picture), your credit balance and how much of the month's free allowance you've used, and a record of each tool run: date, tool, template or journal, outcome, score, and the name and size of the file you submitted. Papers you choose to save to your Library are kept too. Nothing from inside your documents is ever stored.
Only these, and only when you do the thing that sends them. A formatting conversion sends the .docx itself. Submission Letters sends the title, abstract and journal name you pasted (and, for a resubmission or transfer letter, your list of changes and the earlier journal), because a letter has to be written from something. Literature Discovery, Datasets and Find a Journal send your search words or your abstract to look things up. Paper to Slides sends the paper you upload. Figure Builder sends your table's column names and types for its suggestions, never the values; for a diagram it sends the text you paste, to find its structure, and keeps none of it. Reference Check sends the references it reads from your file and nothing else from it. Readiness Check sends nothing from the file at all, with one exception you switch on yourself: a readiness check can confirm the funders in your funding statement against the Crossref Funder Registry and any trial number against ClinicalTrials.gov, and when it does, those names and numbers go and nothing else does. That switch is off until you move it, and the report lists exactly what was sent.
Almost nothing, and the exceptions are things you asked it to keep. Nothing from inside a document is ever stored — not your manuscript, not your abstract, not a drafted cover letter. What persists is your account, your credit balance, and a line per tool run recording the date, the journal or template, the outcome and the file's name and size. If you save a paper to your Library, that is storage you chose: the paper's title, authors, year, journal, DOI and link, plus any note you write, kept until you delete it. If you turn on weekly alerts for a collection, ORBIT checks each week for new papers citing it and puts them in your notifications.
Both, and you don't have to ask us. Account & data in your dashboard downloads everything above as a JSON file, and deletes your account outright — that erases your history and any remaining credits, including paid ones, immediately and permanently. To correct something or raise a grievance instead, email adisingh.cs@gmail.com; we acknowledge within 48 hours.