Script Length Caps, and Why 3,333 English Characters Count as 1,000

Paste an English script on the home page: the counter shows both the raw character count and the cap, already converted.
Where the cap shows
Under the script box on the home page a counter shows characters against the cap. It reads the language of what you pasted: mostly alphabetic text gets the lighter count, mostly Chinese text the full count. Mixed text is judged by which script dominates. The same number decides the parse price shown on the button.

The factor, and why it exists
An English sentence needs roughly three times as many characters as the same sentence in Chinese. Charging and capping by raw characters would make an English script three times as expensive as its Chinese twin for the same amount of story, so alphabetic text is scaled: 3,333 characters count as 1,000. The factor applies to every alphabetic script — Latin, Cyrillic, Greek, Arabic, Hebrew, Armenian, Georgian — and not to Chinese characters, kana or hangul, which stay one-to-one. SceneMixer’s cap, price tag and actual charge all use the same conversion, so the number you see is the number you pay.
The three caps
Scripts under 10,000 equivalent characters are read in one pass. Longer scripts are read in two passes up to 100,000 equivalent characters (for alphabetic text that is 300,000 raw characters). Anything beyond is refused before you pay. How the two-pass read works: Long Scripts Are Read in Two Passes.
What the count prices
Parsing costs 4 credits per 1,000 equivalent characters of the whole script. Each episode’s storyboard costs 1 credit per 100 equivalent characters of that episode’s share of the script. AI Create prices its target length the same way, in the language you write in. Where your credits go in one episode.
AI Create and the cap
AI Create writes to a target length you choose; targets that would not fit the parse cap are hidden from the menu. If a script still comes out over the cap, SceneMixer fails the run, the credits are refunded and no project is created — pick a shorter target and run again. Choosing the Script Length and What It Costs.
| Text | How it counts | Example |
|---|---|---|
| Chinese, Japanese, Korean | 1 character = 1 | 10,000 characters = 10,000 |
| English, German, Spanish, Russian, Arabic… | 3,333 characters = 1,000 | 33,330 characters = 10,000 |
| Mixed | By the dominant script | A Chinese script with English names counts as Chinese |
FAQ
Is the cap in words or characters?
Characters, converted to Chinese-equivalent. For English, divide the character count by about 3.3.
Why is my Japanese script priced like Chinese?
Kana and kanji carry as much per character as Chinese, so SceneMixer counts them one-to-one.
Does the factor change the price after parsing?
No. The cap, displayed price and charge on SceneMixer use the same conversion.
My novel is 400,000 characters. Can I parse it?
Not as one SceneMixer project. Split it into parts under the two-pass cap and parse each part.
Where is the count for AI Create?
In the length menu on SceneMixer's AI Create tab; targets over the cap are not offered.
Paste it and read the counter
The counter already knows your language and the cap; if it is green, you are fine.
Updated November 15, 2026