Synthesia Starter's 12 minutes vs twelve 60-second Sume avatar jobs

Synthesia's pricing page lists Starter at $29 a month for about 12 video minutes and Pro at $89 for about 60, which is $2.42 and $1.48 a minute if you use them all. On Sume there is no monthly bundle: a minute of presenter video is up to one Avatar 1.0 talking-video job, because each job's estimated length must be 4 to 60 seconds. Twelve minutes is twelve jobs, and you pay per job by quality tier.

Synthesia figures are from Synthesia pricing, read 2026-10-06; Sume's from Generate avatar video, read the same day. This post does not put a Sume dollar figure on a minute because the price depends on the quality tier; the live catalog and the cost-by-tier post are the places for that.

How do the plans line up?

What do you get in a minute on each side?

Synthesia lists 9 avatars on Basic, 125 or more on Starter, 180 or more on Pro and 240 or more on Enterprise, with personal avatars from Pro (5) and 160 or more languages and voices, and Full HD downloads on paid plans. It is a studio product with a plan around it.

Sume's avatar flow is an API. You create an avatar once from a prompt, a profile or a photo with POST /v1/avatar-1.0/generate, then call POST /v1/avatar-1.0/talking-video with avatar_handle and a script, or ordered video_inputs for scenes and silence beats. Aspect ratios are 1:1, 3:4, 9:16, 4:3 and 16:9; resolution is 720p. Inline captions can be burned in at the end. Resolution is the plain gap: Synthesia lists Full HD, Sume's avatar video is 720p at this time.

When should you pick which?

Pick a studio plan when people on your team build videos in an editor, you want a library of stock presenters, and your volume is steady. Pick the API when videos are produced by code from a script, a CRM row or an agent, when volume swings, or when you need to approve a first frame before a render, which the preview endpoints support.

Longer than a minute is the one structural limit on Sume's side. Split a six-minute script into six jobs and join them; the 60-second post shows the join. For a head-to-head on features rather than price, read Sume vs Synthesia. Whichever side you choose, run the same 30-second script through both and compare lip accuracy, voice, and how many minutes a revision burns, because revisions are what empty a bundle fastest.

What counts as a minute, and what burns it?

Synthesia's minutes are a monthly allowance, and its yearly plans are priced per month with a minutes-per-year figure: Starter at $18 a month with about 150 minutes a year and Pro at $64 a month with about 720 a year, per the same page. If you buy annually, divide the yearly minutes by twelve before you compare: 150 a year is about 12.5 a month and 720 a year is 60 a month, so the allowances match the monthly plans and only the price falls.

On Sume the unit is the job. A failed job is refunded, a standalone caption pass on a finished clip is a separate job (inline captions on the avatar job are not billed as one), and a re-render of the same script is a new full-price job. That makes draft-then-final cheap to reason about: render at standard for review, then max for the cut you ship, or use a preview to approve stills before any full render. Count your expected revisions per video, multiply by the job price at the tier you use for drafts, and add that to the final render. That one line item is usually larger than the difference between any two rate cards.