ElevenLabs API cost for a 1-minute video: v4, Turbo, music, SFX

Using ElevenLabs' API price page, read 2026-10-03, a one minute video with 900 characters of narration, one minute of music and 30 seconds of sound effects costs about $0.0198 for narration on v4, $0.15 for music and $0.06 for effects, so $0.2298 in total. The 900 character script length and the 30 second effects bed are assumptions for the example, not vendor figures.

The price list

v4 is shown at a 72 percent discount that runs until October 12, so the v4 rate in the table is a promotional rate with an end date.

The worked example

Assume a 60 second video, 900 characters of narration, a full minute of music and 30 seconds of effects.

What dominates

Music and effects, not speech, make up most of the example. Cutting narration cost by half with Turbo saves under a cent. Dropping music saves fifteen cents. If you budget a catalog of videos, model the music line first.

Once the promotional v4 discount ends on October 12, the narration line will change; re-read the page then.

Scaling the example

For 100 videos of the same shape, the example comes to $22.98 in total, of which narration on v4 is $1.98. Music is $15.00 and effects $6.00, which shows again where the money goes. Multiply your own script length, music length and effects length, not these assumptions.

Re-run the sheet after October 12, when the v4 promotional rate ends, and keep a dated copy of the price page.

The same video on Sume

Sume does not mirror these units. Music goes through the Music Router, where every model charges one fixed Music price per audio generation, and the Music Router guide says the catalog lists provider list prices for reference. Speech and transcription are separate: video inspect documents Sume STT 1.0 at $0.01 per audio minute, to be confirmed live in GET /v1/catalog. Read the current TTS rate from the catalog as well, because the docs reviewed here do not state one.

Compare per finished minute, not per unit: sum every audio line your video needs and divide by output length.