# Synthesia for Employee Training: What L&D Teams Get by Plan

> Synthesia for employee training: LMS embeds on every plan, branching from Pro, SCORM and quizzes only on Enterprise, plus Roleplay seats and moderation.

_Source: https://professionalstoolkit.com/articles/synthesia-for-employee-training — The Professional's Toolkit · updated November 2026_

---

Synthesia works for employee training when your team writes scripts and wants a presenter video it can fix by editing text instead of reshooting. The catch is the plan line. LMS embeds, the AI Screen Recorder and PowerPoint import are on every plan, including the free Basic plan, though Basic videos carry a Synthesia watermark. MP4 download and watermark-free videos start on Starter, which is $19 a month billed monthly or $14 a month billed yearly. Branching and CTA buttons start on Pro, which is $89 a month billed monthly or $59 a month billed yearly.[1] SCORM export, quizzes, 1-Click Translation and named-learner video analytics are Enterprise only.

Our verdict: a lean L&D team that only embeds videos into course pages can live on Starter, and Pro earns its price when you need branching, password-protected videos or more monthly credits. A team that needs SCORM tracking is buying Enterprise.


**Enterprise only** — SCORM 1.2/2004 export, quizzes and named-learner video analytics in Synthesia



- **Every plan** — LMS embeds through the Synthesia player (Basic videos carry a watermark)
- **Pro and up** — CTA buttons and branching in training videos
- **$29 a month** — Roleplay learner seat billed monthly ($25 a month billed annually)


## Monday morning: a policy deck becomes a draft module

Most training videos start as something that already exists: a slide deck, a policy PDF, a help article. PowerPoint import and the AI Screen Recorder work on every plan. Starting from a prompt, a pasted script or a document runs through Assistant, the chat-based builder launched in July 2026, which needs Starter or above. Assistant accepts PDF, Word, PowerPoint and text files up to 50 MB each and shows an editable outline before it builds anything. One distinction matters for compliance copy. **Import script keeps your wording exactly**, while Import a file writes a new script from the document. Paste legal-approved text.

Software training runs through the screen recorder, a browser extension for desktop Chrome and Edge that transcribes your voiceover into an editable script. Recordings can last up to 60 minutes, though Synthesia recommends staying under 30, and the recorder has no resolution setting. Recordings made in Camtasia or Loom also work: an MP4 upload of up to 500 MB and 30 minutes turns into editable scenes on all plans. Synthesia's case study says Zoom produced 200+ micro videos in about 6 months with one designer, at 90% time savings.

Generation is where the budget goes. On self-serve plans an avatar video uses about 100 credits per finished minute, charged only on new avatar footage, and previews cost nothing. By our arithmetic from the pricing matrix, Starter covers the 10 included minutes plus up to about 12 more a month, and Pro covers the 10 included minutes plus up to about 60 more. Credits reset every billing period and do not roll over, as our [Synthesia pricing guide](/articles/synthesia-pricing) explains. Plan a pilot module and measure.


> 💡 **Use Import script or paste approved text; don't import the file:** Import script keeps your wording exactly, while Import a file writes a new script from the document. Paste legal-approved policy text.




> ⚠ **Download before you cancel:** After you cancel, videos become inaccessible at the end of the billing period, so download them first; MP4 download needs Starter or above.

## Publishing to the LMS: embed, SCORM and what each one tracks

Synthesia lists about 60 integrations, and most are embeds. Docebo, Articulate 360, SAP Litmos and 360Learning all take the Synthesia embed code once video sharing is switched on, and embeds work on every plan, with the Synthesia watermark on Basic. The share page is not searchable and reaches only people who have the link. Because the course page shows Synthesia's hosted player, a newly published version replaces the old one without re-embedding. There is a flip side. Deleting the public link takes the video down everywhere it appears, SCORM packages included.

SCORM export, in SCORM 1.2 or 2004, is Enterprise only.[2] The package stays under 100 KB because it wraps the hosted player rather than an MP4 file, you choose the completion percentage that counts as finished, and you can switch on learner identity recording. Neither SCORM nor an ordinary embed resumes where a learner stopped; the video restarts from the beginning. Videos with buttons or quizzes cannot be downloaded as MP4 at all. They play only in the Synthesia player.

| Training need | Lowest plan that has it |
|---|---|
| Embed in an LMS or intranet page | Basic (every plan; Basic videos carry a Synthesia watermark) |
| MP4 download without the watermark | Starter |
| CTA buttons and branching | Pro |
| Password-protected videos (help center) | Pro |
| Quizzes with pass thresholds | Enterprise |
| SCORM 1.2 or 2004 export | Enterprise |
| Named-learner video analytics | Enterprise |
| 1-Click Translation and multilingual player | Enterprise |

Video analytics report views, unique views, watch time and average watch time, but only for videos played through the Synthesia player on the share page, an embed or a SCORM package. An MP4 uploaded to another platform reports nothing back. The figures are aggregate and anonymous: naming learners needs Record learner identity in a SCORM package or Require SSO, and both are Enterprise features.[3] Quizzes, with pass thresholds, retries and pass-rate analytics, are Enterprise too. Synthesia's case study says International SOS raised compliance completion by 12% to 97% within two months.

## When the policy changes: one script edit, then every language

Updating is the strongest part. Change the script and regenerate, and Smart Updates push the new version to every link and embed without re-sharing; Replace in version history does the same for an earlier version you pick. Only the changed footage uses credits, so correcting one sentence costs credits for those seconds alone. One gap matters if auditors ask questions. Version history does not log who made each specific edit, only which collaborators were editing when a snapshot was taken.

Translation splits by plan. AI dubbing re-voices any finished video into 70+ languages on self-serve plans and 140+ on Enterprise, where dubbing is a paid add-on an admin enables per workspace. It starts with a clone of the speaker's own voice, but it translates spoken audio only, not on-screen text or burned-in captions. Lip-sync doubles the dubbing credits. 1-Click Translation, Enterprise only, translates the script and on-screen text into 140+ languages, keeps the avatar and layout, and feeds the multilingual player and multi-language SCORM. Translated text can run 20-40% longer than English. Details are in our [video translation guide](/articles/synthesia-video-translation).

Synthesia's case study says Carlisle trains 10,000+ employees on security policies in 7 languages, and that a phishing video took 2 days instead of 3 weeks with an outside producer. That is the pattern the product suits: short policy videos, updated often, in several languages. Team structure shapes the plan as well. Starter and Pro each have a single editor, and colleagues join only as guests who view and comment; workspaces, live collaboration and SSO are Enterprise.


> 💡 **Leave room for translated text:** Translated text can run 20-40% longer than English, so leave space in text boxes before you dub or translate a module.


## Roleplay practice, and the rule about judging people

Roleplay Sessions, launched in October 2026, let a learner hold a live spoken conversation with an Interactive Avatar that asks follow-ups and pushes back; an AI coach then asks the learner to reflect and scores the attempt against a skill rubric. Types include manager feedback and customer conversations, and most sessions take 5 to 10 minutes. Every plan can build roleplays for free. Plays are capped at 10 a month on Basic, Starter and Pro and 25 a month per organization on Enterprise, with no Learner Portal, assignment or tracking. Roleplays use no video credits.

Tracking needs learner seats, a separate subscription at $29 per learner seat a month billed monthly, or $25 a month billed annually. Teams can buy 1 to 100 seats by card without any paid Studio plan. Seats add the Learner Portal, assignment, analytics per learner and uncapped plays for seated learners, while SSO, SCIM (beta) and HRIS sync sit in the sales-led Custom tier. Two limits matter for reporting. SCORM delivers only completion, overall score and pass or fail, without the skill breakdown, and Insights need attempts from at least 10 seated learners.

Those scores come with a hard rule. Synthesia's Acceptable Use Policy prohibits using its outputs as a basis for employment decisions, such as recruitment, promotion, termination, pay or monitoring work performance, and for decisions on access to or evaluation of education and training.[4] Clearly labeled internal training and role-play are not restricted. So a roleplay score works as coaching feedback for the learner and the coach. Never treat it as a rating.

## Sensitive topics, health data and the plan to buy

Every video passes automated and human moderation when it is generated. Stock avatars are limited to neutral, factual content, so medical, legal, financial or political material, or references to real people and companies, can make a script ineligible for a stock avatar. Some of those topics need a Personal or Studio avatar, and some need an Enterprise plan. Only paid plans can appeal a rejection, once per decision, with a review within 48 hours. Earlier approval guarantees nothing. Test your real topics early.

A Personal Avatar of a trainer or leader can unlock some sensitive topics that stock avatars cannot present (some still need Enterprise), with 3 on Starter and 5 on Pro. Synthesia creates a personal avatar or voice clone only with a live consent recording from that same person, who must be of legal age; it refuses celebrities, public figures, deceased people and anyone who has not consented, and you cannot create a photo avatar of someone else even with their permission. Healthcare teams face one more limit. Synthesia is not a HIPAA business associate, so protected health information stays out of scripts.


> ⚠ **Deleting a link breaks every copy:** Deleting a video's public link takes it down everywhere, including LMS embeds and SCORM packages, until the link is recreated.


For most training teams the real choice is a self-serve plan against Enterprise. Starter is enough if you only embed videos; Pro is the step for branching or CTA buttons, password-protected videos or more monthly credits (about 60 extra minutes against about 12, by our arithmetic).

Pro billed yearly at $708 a year saves $360 a year against paying $89 a month, by our arithmetic. At list price the yearly term pays off if you would stay 8 months or more on Pro (9 on Starter); with the first-month offer for new buyers, from 9 months on Pro and 10 on Starter. Purchases are non-refundable even when unused, unless a contract says otherwise, and Synthesia's Customer Terms ask for 30 days' notice of non-renewal. Choose Enterprise only when SCORM, quizzes, translation or named-learner reporting are requirements rather than wishes. Its price is a quote from sales.


## FAQ

**Do closed captions work when a Synthesia video is embedded in an LMS?**

Yes. Closed captions are generated automatically from the script, in the script's language, and viewers toggle them in the Synthesia player, embeds and SCORM exports; they are not embedded in downloaded MP4 files unless you turn on burned-in captions before generating.

**Where does Synthesia store training content?**

Synthesia says customer data is stored in the EU (AWS in Ireland, with backups in Frankfurt) and that it holds SOC 2 Type II and ISO/IEC 27001 certifications.

**Does Synthesia train its AI on our training videos?**

Synthesia says it does not use customer data to pre-train its AI models; fine-tuning on customer data happens only on the customer's written instructions.

**What happens to modules that use older stock voices?**

65 older stock voices retire in December 2026, and videos that use them cannot be regenerated until the voice is replaced, so check voices on modules you update often.


## The bottom line

Synthesia suits L&D teams whose training is script-led, policy-heavy and multilingual, and whose real pain is keeping videos current. Below Enterprise it is a video source you embed: presenter video, cheap script-level updates and aggregate analytics in its own player. The learning-system parts, SCORM, quizzes, named-learner video analytics and one-click translation, all sit on Enterprise, so check which of them your LMS already handles before you price an upgrade. Run your sensitive topics through moderation first, then pilot one module on Starter if you only embed, or on Pro if you need branching or more credits.



## Sources

[1] Synthesia pricing (US): plans, credits and plan comparison — https://www.synthesia.io/pricing (2026-11)
[2] Synthesia docs: Publishing — https://docs.synthesia.io/docs/video-publishing (2026-11)
[3] Synthesia Help Center: How can I view and understand video analytics in Synthesia? — https://help.synthesia.io/en/articles/8271938-how-can-i-view-and-understand-video-analytics-in-synthesia (2026-11)
[4] Synthesia Acceptable Use Policy — https://www.synthesia.io/legal/acceptable-use-policy (2026-11)
[5] Synthesia Help Center: refund and billing policy — https://help.synthesia.io/en/articles/15972602-what-is-synthesia-s-refund-and-billing-policy
[6] Synthesia Customer Terms of Service — https://www.synthesia.io/legal/customer-terms-of-service
[7] Synthesia product updates — https://www.synthesia.io/updates