Turn LaTeX, PDF & Word papers into narrated research video
TeX2Vid and Docs2Vid turn your paper into video — from LaTeX source, or from PDF, Word and Markdown. ScholarReels is the community index where research video is found and kept — helping researchers, educators, and labs expand the reach and impact of their work.
TeX2Vid — LaTeX to Video
Let your research reach more people. Convert your LaTeX paper into a grounded social-media teaser, a short, or a video abstract — that researchers, students, and conference audiences can watch.
Upload a .tex file or archive and get back a narrated MP4, the Beamer slides as a PDF, subtitles and the narration script — automatically. Highlights and Video Abstracts also hand back the editable Beamer .tex sources.
- Research Teasers (30 s) and Shorts (50 s) for social media; Highlights (1 min) and Video Abstracts (3 min) for conferences and journals.
- Multiple Beamer themes (Madrid, CambridgeUS, Singapore, Rochester).
- Every paid tier: editable slides, swappable figures, and your own voice — or the default AI narration.
- Figures and results sourced directly from your paper.
Docs2Vid — Docx/PDF to Video
The same grounded pipeline, without LaTeX. Convert a PDF, Word or Markdown paper into a grounded teaser, short, highlight or video abstract — or hand it an arXiv link and let it fetch the paper for you.
Docs2Vid automatically extracts the title, authors and affiliations, the figures, and the paper’s important sections — abstract through conclusion — gives you a moment to correct the metadata, then renders a narrated MP4 — with a PowerPoint deck on the 16:9 formats.
- Research Teasers (25 s) and Shorts (50 s) for social media; Highlights (1 min) and Video Abstracts (3 min) for conferences and journals.
- PDF, Word (.docx) or Markdown, up to 50 MB — or an arXiv link or ID.
- Every paid tier: editable slides, swappable figures, and your own voice — or the default AI narration.
- Free weekly Research Teaser — no signup.
ScholarReels — the scholarly index and repository for research videos
The scholarly index and repository for high-quality research videos — a community index of Shorts, Highlights, and Video Abstracts, with every claim traceable to the source paper.
Watching is free and open — no paywall, no login to browse. Videos are citable, embeddable, and discoverable, and nothing is auto-published: a moderator checks every video.
- Research Shorts (20–90 s, 9:16), Highlights (20 s–2 min, 16:9), and author Video Abstracts (over 2–5 min, 16:9) — duration limits strictly enforced.
- Every video vouched for — a machine grounding review for tool-made videos, or peer attestation by the author or a PhD-credentialed reviewer.
- A living index — subscribe to subjects, authors, venues, and institutions; verified channels for conferences, journals, and institutions.
- Open access — free to watch, citable, and embeddable; DOIs for author-attested Video Abstracts coming soon.
Want to be a reviewer, or host your organization’s research videos on a verified channel? Contact us.
Built for research papers from most fields
TeX2Vid works with LaTeX templates from journals and conferences like
Make a video abstract from your paper
A video abstract is a short video — typically around three minutes — in which the authors present a paper’s problem, motivation, key results and conclusion. Journals and conferences increasingly invite one alongside the manuscript, and papers with a video abstract are easier to find, skim and share.
Paper-to-video tools generate video abstracts from the research papers you upload: upload LaTeX source to TeX2Vid, or a PDF, Word or Markdown file to Docs2Vid, and get a narrated 3-minute 16:9 Video Abstract — with the slide deck, and every claim checked against the source paper. Then publish it on ScholarReels, where author video abstracts are indexed, citable and open access.
Frequently asked questions
What do I upload, and what do I get back?
For TeX2Vid, upload your LaTeX source — a .tex file, a .zip or .tar.gz archive, or a project folder, up to 50 MB. Every TeX2Vid render, the free Teaser included, comes back with the MP4, the Beamer slides as a PDF, subtitles as .srt and .vtt, and the narration script as a text file. Research Highlights and Video Abstracts add the editable Beamer .tex sources, and ship a 16:9 MP4 alongside a 9:16 cut; Teasers and Research Shorts are 9:16 video. Slides come in 16:9 and 9:16; the downloadable deck is the 16:9 one. With Docs2Vid, you upload a PDF, Word (.docx) or Markdown file — or paste an arXiv link — and get back the narrated MP4, plus a PowerPoint deck on the 16:9 formats.
Do I need an account?
Not for the free Research Teaser — a 30-second, 9:16, SheQAI-branded teaser renders without signing in, subject to a daily quota on a first-come, first-served basis. Docs2Vid’s free Teaser — 25 seconds, weekly — also renders without an account. Paid renders need an account on either.
What do TeX2Vid and Docs2Vid cost?
One credit is $1 USD, and the same credit balance spends on either generator. On both, a Research Short is 8 credits ($8), a Research Highlight 10 credits ($10), and a Video Abstract 20 credits ($20). TeX2Vid also offers an unbranded Research Teaser at 5 credits ($5); Docs2Vid’s Teaser is free-only. A 30-credit pack is $24, a 20% saving on pay-as-you-go. Monthly plans run $18 for 24 credits and $65 for 100 credits, and institutions can arrange volume pricing; all of them cancel anytime. The current numbers are always on the TeX2Vid pricing page and in Docs2Vid’s pricing section.
What formats does TeX2Vid make, and how long are they?
Four. A Research Teaser is 30 seconds in 9:16, and a Research Short is 50 seconds in 9:16 — both vertical, for Shorts, Reels and TikTok. A Research Highlight is 60 seconds in 16:9, and a Video Abstract is three minutes in 16:9, structured as problem, motivation, analysis and conclusion — both horizontal, for conferences and journals. Docs2Vid makes the same four from PDF, Word or Markdown — its Teaser runs about 25 seconds, and its 16:9 formats ship a PowerPoint deck instead of Beamer.
How do I know it hasn’t made something up?
Every render is grounded in your paper and gets an independent faithfulness review against the source. The load-bearing claims are each paired with the span of the paper they come from, in a grounding report you can read before you publish anything.
Can I narrate it in my own voice?
Yes — on both TeX2Vid and Docs2Vid. Renders ship with AI narration by default, and every paid tier includes the edit studio: editable slides, swappable figures, and narration in your own voice.
How long do you keep my videos?
On TeX2Vid, paid renders are retained for 90 days, and free renders for 30. Download the MP4, the slides, the subtitles and the narration script and they are yours to keep for good.
Does my video have to be made with TeX2Vid to appear on ScholarReels?
No. ScholarReels is the scholarly index and repository for research video, and it is open to any researcher and to video made with any tool. Watching is free and open — no paywall, and no login to browse.
What can I post on ScholarReels, and how long can it be?
Three formats, and ScholarReels sets its own lengths — it takes video made with any tool, so its limits are not TeX2Vid’s. Shorts run 20–90 seconds in 9:16, Highlights 20 seconds to 2 minutes in 16:9, and Video Abstracts strictly over 2 minutes and up to 5 in 16:9, those last limited to authors posting accepted or published papers. The durations are strictly enforced. Every TeX2Vid format fits: Teasers and Research Shorts both post as Shorts, Research Highlights as Highlights, and Video Abstracts as Video Abstracts. ScholarReels simply has no separate Teaser category — a 30-second Teaser lands inside the 20–90 second Short window.
How does a video get onto ScholarReels?
You sign in with an institutional email. A TeX2Vid render goes straight to a moderator, carrying its grounding report. A video you made yourself is confirmed through a link emailed to your institutional address, and is backed by a human attestation — your own as the author, or a PhD-holding reviewer’s — before it reaches a moderator. Nothing is auto-published: a moderator reviews every video before it appears.
More tools in exploration. Have a use case you’d like us to tackle? Reach out to us.