Synthesia: Workflow, Credits and Output Review
Script-led avatar video and dubbing, with shared AI credits and distinct sharing, download and branding entitlements.
Who should consider Synthesia
Training and enablement teams turning approved procedures into repeatable explanations. It is best evaluated on clarity, maintainability and review effort rather than presenter realism alone.
Features and pricing
Synthesia lists avatar videos, video assistance, dubbing and translation. Its current pricing page separates video/session access from API pricing and describes shared credits across AI features.
Basic is $0 and lists 500 credits per month plus a free allowance, with watermarked sharing. Starter is $29 billed monthly with 1,250 credits and download/logo-removal features. Credit consumption differs by activity; do not add every displayed maximum together as simultaneous entitlement. API costs have a separate pricing tab.
Official product website · Official plans and limits
A practical evaluation task
Convert a short, approved operating procedure into a two-minute training draft. Include a condition where the viewer must stop and ask a supervisor rather than continue.
Provide the exact procedure version, intended audience and forbidden changes. Review the script before generating media. Keep a mapping from each scene to the source step so the video can be updated when the procedure changes.
What a useful result must pass
Check that the stop condition is spoken and visible, sequence remains correct and narration does not skip exceptions. Reopen the exported file in the learning environment you intend to use. Confirm captions and distribution requirements before purchasing a tier.
Limits and data access
A generated training video needs subject-matter approval. Shared credits can be consumed by activities other than video, and a share link is not the same entitlement as a downloadable file. Treat custom avatars and uploaded staff material as permission-sensitive.
Alternatives and tradeoffs
Compare HeyGen for presenter-led marketing and localization trials. Use Descript when the lesson depends on editing a real demonstration rather than generating an avatar explanation.
Sources checked October 22, 2026. Based on official product documentation. Use the suggested trial to assess fit for your workflow.