Turn an SOP or Compliance Manual into a Narrated Training Video
To turn an SOP, policy, or compliance manual into a narrated training video in Scholarly, upload the PDF (or Word or PowerPoint file), open New → Video lecture, select the file and the pages that matter, choose a length, language, and voice, and press Create. Scholarly writes the narration only from the material you selected and renders an animated, chaptered video with synced captions, a searchable transcript, and built-in question checkpoints. You share it with your team's Library or by link, and on a paid seat you can download a watermark-free MP4. Premium and Laureate seats, including Premium for Teams and Laureate for Teams, can target any length up to 45 minutes and narrate in over 70 languages.
Scholarly makes the training. It is not a learning management system: it does not enroll people, set due dates, track who finished a video, or issue completion certificates. If you need those, keep your LMS and upload the finished MP4 and captions to it. See Scholarly Teams vs an LMS.
What you get
- A narrated, animated video scripted from your document, with one narrator from start to finish.
- Chapters that follow the stages of the procedure, so a viewer can jump to step 4 without rewatching steps 1 to 3.
- Synced captions in the lecture's language, written from the same script as the narration, and a searchable transcript.
- Interactive question checkpoints that slide up during playback, plus an AI chat beside the video that answers from the transcript and your original files.
- On a Standard lecture, a companion slide deck you can download as a PDF.
- On a paid plan, an MP4 download and an export pack: one zip with the video, the slide deck, the narration script by chapter, captions as
.vttand.srt, and a README listing the sources.
There is no avatar, presenter, or lip-synced face. The video is narrated visuals built from your material: step sequences, comparisons, definitions, callouts, charts, and illustrations.
Before you start: get the source right
The video can only be as accurate as the document behind it.
- Use the approved, current version. If there are two versions of a policy in your library, remove the old one first.
- Prefer a text PDF or the original Word file. Scanned manuals work, but accuracy tracks the scan quality.
- Pair the procedure with the policy behind it when they live in separate files. One video can combine up to 10 sources on Premium and up to 20 on Laureate.
- Long manual? Plan one video per section. A 200-page compliance manual makes a vague single video; chapter 4 of it makes a sharp one. Short, dated modules are also easier to refresh when one section changes.
Optional first step: run the manual through AI SOPs to get a clean, numbered procedure PDF, then make the video from that SOP. Steps the SOP inferred rather than read are labelled for review, which makes the source easier to approve before anything is narrated.
Step by step
1. Add the document
Upload it from New → Upload, pick it from your library, choose it from Google Drive, or paste a link to an online PDF. Accepted inputs include PDF, Word (.docx), PowerPoint (.pptx, .ppt), text, Markdown, CSV, and images such as photos of a printed procedure. If you are on a team, put the source in a team folder or team project first so everyone works from the same copy (see step 8).
2. Open the video lecture creator
Open New and choose Video lecture. You can also start from the file itself: after an upload, open the More menu and choose the video lecture option. Or ask AI Chat: "Turn this SOP into a training video."
3. Select the source and the pages
Pick the file. For a long PDF, use the page selector to choose only the section this video should teach. Scholarly reads the selected pages, including tables, diagrams, and annotations, before it writes anything.
4. Decide how the narration is written
Two options:
- Teaching script (default). Scholarly plans a lesson from your document: what a new employee must understand first, which steps deserve their own scene, where a caution belongs. The wording is new, the facts come from your file.
- Use my script as written. Turn this on when the wording itself is approved, for example a compliance statement your legal team signed off. Scholarly narrates your source word for word instead of writing a new script. The length follows your script within the length you choose, so a long script may be shortened to fit a short target. Pick a longer target for a longer script.
5. Choose the settings
| Setting | What to pick for training |
|---|---|
| Length | 2 min, 15 min, 45 min, or Custom (1 to 45 minutes) on a paid plan. Free lectures target about 2 minutes. |
| Video Mode | Standard for a clear, predictable editorial lesson; Freestyle (the paid default) lets the AI pick the format. See Video Modes. |
| Language | Any of over 70 output languages, or Auto to match the source. The source and the video do not have to share a language. |
| Voice | Auto, or pick a named narrator after hearing its preview. |
| Video quality | 720p by default; 1080p Full HD on paid plans. |
| Scholarly logo | On by default. Paid plans can turn it off. Turn it off before generating if you need a watermark-free file, because a logo cannot be removed later. |
| Customization Profile | Apply your team's shared profile so every training video uses the same instructions, colors, and logo. |
| Custom instructions | For example: "Audience is new warehouse staff, plain language", "Call out every safety step", "Use our tool names exactly as written." |
6. Create
The Create button shows the total cost in AI Creation Credits before you start, counting both the model and the length. Generation runs in the background, so you can close the window. Short lectures often take 10 to 25 minutes; a long target can take more than an hour, especially at 1080p. A live progress page shows how many scenes are planned and recorded.
If a creation fails, its AI Creation Credit is returned automatically and the failure screen offers the fix that applies, such as picking fewer pages or replacing the source.
7. Review before you share
Treat a generated module as a first draft. Before it reaches every new hire:
- Read the transcript against the source. It is faster than watching, and it is word for word what the narrator says.
- Check the requested length against the delivered runtime. The video page prints both, plus the pages the lecture was built from and its language.
- Ask the video's AI chat where a specific claim came from. It answers from the transcript and your original files.
- Fix by regenerating, not by hand. There is no scene-by-scene editor. To change what the video says, correct the source, narrow the pages, adjust the custom instructions, or switch on Use my script as written, then generate again. You can rename a finished video.
8. Share it with your team
- Your team's Library. Open Share, and under Who can open it choose Everyone at your team's name. Every active member can open it from Team → Library, where it shows who shared it and when, and a view count once people open it. It also appears in each member's All Content with a Team badge.
- A team-owned project. Keep the source, the video, the SOP, and the quiz together in one team project, for example "Onboarding: Warehouse safety". Everything stays with the team if the person who built it leaves.
- A link. Under Who can open it, choose Anyone with the link (it is not listed in Explore or search; avoid Public, which is) and send the link. Viewers watch in the browser, read the transcript, and browse chapters without an account. You can add a password, an expiry, or revoke the link later. See Password, Expiry, and Revoking a Shared Link.
- A file. On a paid plan, Download video saves the MP4, and Download export pack adds the slides, script, and
.vttand.srtcaptions for an intranet or LMS upload.
9. Add a knowledge check
Every video already includes question checkpoints. For a separate check, generate a quiz, exam, or flashcard set from the same source file, or create flashcards straight from the finished video. One source, one lesson, one check.
Length, plan, and credit limits
| Free | Premium (and Premium for Teams) | Laureate (and Laureate for Teams) | |
|---|---|---|---|
| Video length | About 2 minutes, fixed | 15 minutes by default; any target up to 45 minutes | Same as Premium |
| Sources per video | 1 | Up to 10 | Up to 20 |
| PDF pages read | First 32 pages | Up to 1,000 pages | Up to 3,000 pages |
| Upload size | 8 MB | 300 MB | 1 GB |
| Resolution | 720p | 720p or 1080p | 720p or 1080p |
| Scholarly logo | Always on | Can be turned off | Can be turned off |
| MP4 download and export pack | No (view link only) | Yes | Yes |
| AI Creation Credits | 1 for life | 15 each week | 80 each week |
Credit cost of one video: the model's credits plus the length. On Auto (GPT 6 Luna) a video costs 1 credit. A target over 15 minutes adds 1 credit, and a target over 30 minutes adds 2. So on Auto, a 15-minute module costs 1 credit, a 20-minute module 2, and a 45-minute module 3. Other models cost more: Kimi K3 and Claude Sonnet 5.5 cost 2, GPT 6.1 Sol, GPT 6 Astra, and Claude Opus 5.5 cost 3, and Gemini 4 Argon and Muse Wonder cost 4, before the length credits. If the finished video lands in a shorter length tier, the unused length credits are returned.
Weekly credits reset every Monday at 00:00 UTC and belong to each seat, so one person's videos never use up a teammate's allowance. On a team, an Owner or Admin can buy a member a one-time credit pack on the team bill. In some countries, video lectures are not available on the Free plan.
Languages
Narration, on-screen text, captions, title, and chapter headings follow the language you choose, from over 70 output languages including Spanish, Portuguese, French, German, Arabic, Hindi, Japanese, and Chinese. Arabic, Hebrew, and other right-to-left lectures are laid out right to left. An English SOP can become a Spanish or Arabic training video: generate it once per language your team needs. See Languages in Scholarly.
What Scholarly does not do
Be clear about these before you plan a rollout around it:
- No avatar or on-screen presenter. If a talking head is a requirement, an avatar video tool is the right purchase.
- Not an LMS. No enrollment, assignments, due dates, or completion certificates.
- No per-person completion tracking. Team Activity shows counts per member (creations by type, AI generations, chat messages, study sessions), and Library items show a view count. It does not record which video each person finished or whether they passed.
- No SCORM or xAPI package. For an LMS, upload the MP4 with its
.vttor.srtcaptions as a video asset. - No timeline editor. Change the source or instructions and regenerate.
- Not a substitute for review. Grounding means every claim should trace to your document, but a compliance module still deserves one pass from someone who knows the process.
Facts at a glance
- Inputs: PDF, Word, PowerPoint, text, Markdown, CSV, images, Google Drive files, web or PDF links, or a typed prompt.
- Output: narrated, animated video with chapters, synced captions, searchable transcript, question checkpoints, AI chat, and (Standard mode) a slide deck.
- Maximum length: 45 minutes on paid plans; about 2 minutes on Free.
- Sources per video: 1 on Free, 10 on Premium, 20 on Laureate.
- Languages: over 70 output languages, right-to-left layout for Arabic and Hebrew.
- Downloads: MP4 and export pack (video, slides, script,
.vttand.srtcaptions) on paid plans; a view link on every plan. - Cost: 1 AI Creation Credit on Auto up to 15 minutes; +1 above 15 minutes, +2 above 30 minutes.
- Team pricing: Premium for Teams $35 per seat per month or $336 per seat per year; Laureate for Teams $110 per seat per month or $1,056 per seat per year; self-serve for 1 to 29 seats.
- Not included: avatars, SCORM, enrollment, due dates, completion tracking, certificates.
Frequently asked questions
Can I turn a 200-page compliance manual into one video?
You can select it, but you will get a better result from one video per section. Use the page selector to pick each section, and keep the modules in one team project so they stay together.
Will the video say anything that is not in my document?
The narration is scripted from the pages you selected, not from general web content. Review the transcript before rollout anyway. If the exact wording matters, turn on Use my script as written so the source is narrated word for word.
How do new employees watch it?
Through the team Library, a team project, or a link. Viewers do not need an account to watch a link set to Anyone with the link. On a paid plan you can also download the MP4 for your intranet or LMS.
Can I see who watched it?
You can see a view count on Library items and per-member activity counts in Team → Overview → Activity (Owners and Admins). Scholarly does not record per-person completion of a specific video. If you need completion records or attestations, host the MP4 in your LMS.
Does it export SCORM?
No. The export pack gives you the MP4, slides, script, and .vtt and .srt captions, which most LMSs accept as a video asset.
What happens when the SOP changes?
Replace the source document and generate the video again, then swap the link or the Library item. Regenerating keeps the training in step with the document instead of maintaining a second, hand-edited version.
Can I make the same video in Spanish?
Yes. Generate it again with Language set to Spanish. The source can stay in English.
Is there an avatar or presenter?
No. Scholarly produces narrated visual lessons without a synthetic presenter.
Related
- AI Video Lectures: every setting and player feature in full.
- AI SOPs: turn a manual into a numbered procedure first.
- Set Up a Shared Training Library for Your Team: roles, invites, the Library, and Activity.
- Scholarly Teams vs an LMS: when you need which, and how to use both.
- Exporting Your Work: the export pack and other formats.
- Turning Work Documents into Training Material: recipes for onboarding, enablement, and expert capture.