From File to Narrated Video: Turn a Document into a Talking Presentation with AI (2026)

Quick Answer (≤60 words): Almost every AI presentation tool stops at a static deck - you still have to present it live. 2Slides goes further: upload a file, it builds the slides, generates per-page AI narration (single- or multi-speaker), and exports an H.264 MP4. A report becomes a watchable video your audience can play on their own time.

This article is for anyone who needs a presentation that plays itself: an onboarding deck, a product update, a course module, a sales walkthrough, an investor report that has to reach people who will never sit in a live meeting. You have the source material - you want a finished video, not just slides.

We are 2Slides. This workflow is a genuine capability gap in the market, so we will be specific about how it works and where it does not apply.

---

Key takeaways

• A deck is not a presentation. Most AI tools generate slides and stop; the narration and video - the part that lets an absent audience actually consume it - is left to you.
• 2Slides runs file → slides → narration → video in one flow. Upload a document, generate the deck, add per-page voiceover, and export an MP4 - without leaving the tool.
• Narration is per-page and multi-speaker capable. Each slide gets its own generated voiceover; you can use a single narrator or assign multiple speakers. Cost is 210 credits per page (10 for the script, 200 for the audio).
• Video is real 16:9 or 9:16 MP4. H.264 at 1920×1080 or 1080×1920 - landscape for desktop, vertical for social and mobile.
• It only works on the right job type. Narration and video apply to file-based and "create like this" decks - not the template-only Fast PPT path.

---

Why "AI presentation" usually means "AI slides, and then it's on you"

The headline tools of 2026 - the ones that crossed $100M ARR and a $2.1B valuation in Gamma's case - are excellent at one thing: turning a prompt into a good-looking deck. But a deck still assumes a presenter. For the large and growing share of communication that happens asynchronously - the teammate in another timezone, the prospect who skims on their phone, the new hire onboarding at 11pm - a static deck is only half the deliverable.

Adding voiceover and rendering a video has traditionally meant a second tool and a manual pipeline: export slides, record audio track by track, line it up in a video editor, encode. That friction is why most "AI presentations" never become videos at all.

---

The 2Slides workflow: file → slides → narration → video

2Slides keeps the whole chain in one place, starting from a file.

1. Turn your file into slides. Upload a report, PDF, or data file and let Create from File build the deck - the same workflow covered in Spreadsheet to Slides and PDF to Slides. (You can also start from a reference design with "create like this.")
2. Generate per-page narration. 2Slides writes a script for each slide and synthesizes voiceover - single narrator, or multiple speakers for a dialogue feel. Narration runs per page, so the audio matches what is on screen.
3. Export the video. Render an H.264 MP4 at 1920×1080 (16:9) or 1080×1920 (9:16). The result is a self-contained video file you can upload anywhere.

Because all three steps share the same job, the slides, the audio, and the video stay in sync - no manual re-alignment.

---

What the narration actually does

The narration step is not a generic text-to-speech read of whatever text is on the slide. 2Slides generates a per-page script and produces audio for each page, so the voiceover explains the slide rather than just reading its bullet points aloud.

• Single-speaker for a straightforward briefing or course module.
• Multi-speaker when you want a two-voice, podcast-style delivery.

Each page is priced at 210 credits for narration - 10 credits to generate the script and 200 to synthesize the audio - so a 10-slide deck narrates for 2,100 credits. These figures come from the live narration pipeline.

---

Landscape or vertical: one deck, two formats

The MP4 export supports both orientations, which matters because the destination dictates the shape:

The same narrated deck can be exported either way, so a quarterly update can ship as a widescreen video for the all-hands and a vertical cut for a mobile-first team.

---

Worked example: an onboarding PDF → a narrated video

You have onboarding-guide.pdf.

1. Upload it to Create from File; generate a 12-slide deck.
2. Add single-speaker narration across all 12 pages: 12 × 210 = 2,520 credits.
3. Export a 1920×1080 MP4 for the LMS, and a 1080×1920 cut for the mobile app.

A new hire now watches a narrated walkthrough instead of reading a PDF - and you built it once, from the file you already had.

---

When NOT to use this

1. You need an editable .pptx, not a video. Narration and video live in the Workspace flow. If your deliverable is a PowerPoint file to hand off, use the template-driven Fast PPT path - it produces editable .pptx and does not include narration.
2. A live presenter is the point. A keynote or a sales call where the human delivery matters does not benefit from pre-rendered narration - present it live and skip the video step.
3. Tightly regulated voice/script requirements. If the spoken script must be legally reviewed word-for-word, generate the narration text first, have it approved, then synthesize.

---

Frequently Asked Questions

Can AI turn my slides into a narrated video?

Yes. In 2Slides, a file-based deck (or a "create like this" deck) can have per-page AI narration generated and then be exported as an H.264 MP4. The result is a self-contained video with synchronized voiceover - not just a slideshow. The template-only Fast PPT path does not include narration.

Does the AI narration just read the slide text aloud?

No. 2Slides generates a per-page script and synthesizes audio for each slide, so the voiceover explains the slide rather than reading its bullets verbatim. You can choose a single narrator or multiple speakers.

What video formats and resolutions are supported?

H.264 MP4 in two orientations: 1920×1080 (16:9 landscape) and 1080×1920 (9:16 vertical). The same narrated deck can be exported in either, so you can target both desktop and mobile/social.

How much does narration cost?

210 credits per page - 10 to generate the script and 200 to synthesize the audio. A 10-slide deck narrates for 2,100 credits. Slide generation and video export are billed separately.

Can I start from a document instead of writing slides manually?

Yes - that is the recommended path. Upload a PDF, report, or data file via Create from File; 2Slides builds the deck, then you add narration and export video. See Spreadsheet to Slides and PDF to Slides for the file-to-deck step in detail.

Is there an API for this?

Slide generation and narration are available through the 2Slides API: the file-to-deck job runs via POST /api/v1/slides/create-pdf-slides, and narration via POST /api/v1/slides/generate-narration for eligible jobs.

---

The bottom line

The market solved "prompt to slides." It mostly did not solve "file to a finished, narrated video." 2Slides runs the whole chain - upload, generate slides, narrate per page, export MP4 - so an async audience gets something they can actually watch.

Try it: start at Create from File, build a deck from a document, then narrate and export. For the slide-building step, see From Spreadsheet to Slides and PDF to Slides.

Related reading: From Spreadsheet to Slides · PDF to Slides · 7 Best Gamma Alternatives in 2026 · AI Presentation API Comparison with Code Examples

Last reviewed: 2026-06-30 by the 2Slides team.