How to make a training video with voiceover, music and captions, no video editor

This page may contain affiliate links from which Blur Studio earns commissions.

Simple training videos, a screen recording with narration, a slide deck turned into something watchable, shouldn't need the same production process as a marketing video. The problem is that most "simple" options still mean exporting a voice file, importing it into separate editing software, timing captions by hand, and finding royalty-free music somewhere else. That's three or four tools for one short video.

What ElevenLabs Studio actually is

ElevenLabs Studio is a browser-based editor that puts voiceover, music, sound effects, captions, and now video and images, on one timeline. As of its September 2026 relaunch (Studio 4.0), it works as a prompt-first editor: you describe roughly what you want and it generates a starting draft you then refine on the timeline, rather than starting from a blank project every time.

Project limits scale with your plan: the free tier includes 500 total Studio projects, rising to 750 on Starter, 1,000 on Creator, 3,000 on Pro, and 20,000 on Scale and Business. Within any single project, you get up to 500 chapters, 400 paragraphs per chapter, and 5,000 characters per paragraph, which is more structure than a short training video needs but matters if you're producing something closer to a full course. Long narration is covered in ElevenLabs for audiobooks and long narration.

Export quality is the thing worth checking before you commit to the free tier for anything client-facing: Free, Starter and Creator export at 128kbps MP3 (or WAV generated from that same 128kbps source), while Pro and above export at 44.1kHz WAV or 192kbps MP3. Video exports on the free plan also carry a watermark, which paid plans remove.

Background music inside Studio comes from ElevenLabs' own Eleven Music, generated to length rather than pulled from a stock library you'd need to license separately, so you aren't searching a royalty-free site and hoping the licensing terms actually cover internal training use. That matters more than it sounds for training content specifically, since a lot of companies reuse training videos for years, hand them to contractors or partners, or post excerpts publicly for recruiting, and a stock track with ambiguous licensing terms is exactly the kind of thing nobody checks until it's a problem. Generated-to-length music sidesteps that question entirely for anything you build inside Studio itself. If you want a stock track instead, see where to get background music without license fees.

A worked example: a 10-minute onboarding module

Say you're building a new-hire onboarding video covering five topics: company overview, where to find key systems, who to contact for what, safety basics, and first-week expectations. Structured as five chapters in one Studio project, each roughly two minutes, that's about 1,500-1,800 words of script total, comfortably inside the per-chapter limits even on a free account. Write each chapter's script separately rather than one long block; Studio's chapter structure exists specifically so you can regenerate or re-time one section (say, the systems-access chapter, which tends to change most often as tools change) without touching the other four. Generate the voiceover chapter by chapter too, and listen to each one before moving on, since catching an awkward phrase in a two-minute chapter is a five-second fix; catching it after you've built captions and timed music around a full 10-minute file means redoing more work to isolate and fix it.

Captions aren't just a nice-to-have for training content

For training specifically, captions do real work beyond accessibility, though accessibility matters too, especially if any new hire has a hearing impairment or if the video will be watched in a noisy warehouse or shop floor where audio isn't practical. Captions also help non-native English speakers follow technical terms and acronyms they might not catch by ear alone, and they make the video usable for someone skimming back through it later to find one specific point without replaying the whole thing. Check Studio's auto-generated caption timing against the actual voiceover rather than trusting it blindly, particularly around acronyms, product names, or anything spelled unusually; auto-captioning tools in general are more likely to mishear an unfamiliar term than a common word, and a wrong caption on a safety instruction or a system name is worse than no caption at all, since it reads as confidently wrong.

How to make the video, step by step

1. Write the script first, ideally broken into the same sections your video will have, since Studio's chapter structure maps naturally onto a multi-section training video.
2. Generate the voiceover through ElevenLabs Studio and listen to the full thing once before adding anything else. Catching an awkward phrase or mispronunciation before you've built captions and timed music around it saves redoing that work.
3. Add captions on the same timeline, checking the timing against the voiceover rather than trusting auto-generated timing blindly, especially around any technical terms or acronyms.
4. Add music at a level that sits under the narration, not competing with it; training content specifically needs the voice to stay clearly intelligible, more so than a marketing video where mood matters more than every word landing. Music and effects options are covered in ElevenLabs sound effects and music for business users.
5. Export at the quality your plan actually supports, and if you're on a paid plan, confirm the video export doesn't carry the free tier's watermark before sending it anywhere official.

How this compares to other training video tools

Tools like Descript or Canva's video editor cover similar ground, timeline editing, AI voice, captions, and each has real strengths: Descript's text-based editing (cut the video by deleting words from a transcript) is genuinely faster for heavy editing than a traditional timeline, and Canva's template library is deeper if visual design matters more than narration quality to your training content. Studio's advantage for training specifically is that the voice generation and the editing timeline are the same product, built by the same company, so there's no export-import step between getting the narration right and placing it on the timeline, and no separate subscription to coordinate. If you're already paying for one of those other tools and it's working, this isn't necessarily a reason to switch; if you're starting from nothing and training video is mainly narration plus simple visuals and captions, starting with Studio means fewer tools to learn and fewer subscriptions to juggle.

What I'd actually tell a training manager trying this

For a single-presenter explainer, a screen-share walkthrough, or a short onboarding module, this genuinely replaces the "export, import, re-sync captions, find music" workflow with one tool. Where it stops being the right tool is complex video work: multi-camera cuts, heavy visual effects, or anything where the editing itself, not the narration, is the hard part. Studio's prompt-first draft is a good starting point for simple structures; it's not pretending to be a full non-linear editor, and treating it like one past its actual scope just means fighting the tool.

When to skip this

If your training video needs real video editing, screen recording cleanup, cutting between multiple camera angles, visual effects, Studio's timeline isn't built for that depth, and a dedicated video editor will get you there faster even with the extra steps of moving audio between tools.

Reviewing and updating training videos as a team

Training content rarely gets made once and left alone; policies change, tools get replaced, and a video that was accurate six months ago can quietly become wrong in one specific chapter while the rest stays fine. This is where Studio's chapter structure earns its keep over just having one long exported video file: if your systems-access chapter changes because you switched tools, you can regenerate and re-time that one chapter and re-export, rather than redoing the full 10-minute video from scratch. Before publishing any update, have someone who didn't write the script watch the whole thing once, not just the changed section, since a fix in one chapter occasionally creates a small continuity gap with what the next chapter assumes the viewer already knows. Keep a simple log of what changed and when, even just a line in a shared doc, so six months from now whoever owns the content knows which chapters are current and which might need another pass. Larger collections are covered in building a training video library without a studio.

How I'd approach this

For producing training videos one at a time, ElevenLabs' Studio covers this well without needing anything custom. Where I get involved is when a business wants to produce training content at volume, a library of courses, generated from existing documentation or SOPs, kept current automatically as the underlying process changes. That's a different problem than making one video, and it's the kind of pipeline I build once a team has outgrown making videos by hand one at a time.