Elai lists 75+ languages: how to cover them with Sume Avatar

Elai's pricing page lists 80+ avatars and 75+ languages. Sume Avatar 1.0 is English only, so for a Spanish or Korean talking shot you generate the voice with TTS 1.0 and then drive a still with a lip-sync model.

Language coverage as each page states it

The Elai figures are the vendor's own. The Sume row describes how a non-English shot is built today, not a language list for Avatar.

Treat the Sume route as a two-model pipeline you assemble, where Elai sells it as one product.

The non-English route on Sume

Step one is the voice: POST /v1/tts-1.0/generate takes up to 20,000 characters and returns audio up to 1,200 seconds. Step two is the face. POST /v1/veed/fabric-1.0 takes a still plus Sume-hosted audio (10 MB audio limit), and POST /v1/minimax/h3-max/lip-sync takes audio of 5 to 14.8 seconds.

Hold your script in the target language yourself; Sume does not translate. Burn captions afterward with the cues field so the on-screen text matches the spoken line exactly.

Captions for Korean

Korean captions need a Hangul style such as black-outline. The Latin styles slam, punch and tiktok-green reject Korean text with a 400 caption_hangul_text_latin_style, so pick the style before you render, not after.