H3 Max lip sync at 14.8 s bills 15 s: $1.50 at 768p, $3.00 at 1080p

A 14.8-second line is the longest MiniMax H3 Max lip sync accepts on Sume. It rounds up to 15 billed seconds: $1.50 at 768p, $3.00 at 1080p, $0.9375 at 480p.

4 min readSume
All posts

A 14.8-second line is the longest audio MiniMax H3 Max lip sync takes on Sume, and it bills as 15 seconds: $1.50 at 768p and $3.00 at 1080p. The route charges ceil(duration_seconds) times the provider list per second times 1.25, and ceil of 14.8 is 15. The extra 0.2 seconds cost $0.02 at 768p.

Sume accepts duration_seconds from 5 to 14.8 and refuses anything outside that range; it never clamps.

The three prices

The list rates on the 2026-09-22 rate card are $0.05, $0.08, and $0.16 per second at 480p, 768p, and 1080p. Sume multiplies each by 1.25.

A 14.8-second line billed as 15 seconds (list as of the 2026-09-22 rate card)
ResolutionList per secondSume per second15 billed seconds
480p$0.05$0.0625$0.9375
768p (default)$0.08$0.10$1.50
1080p$0.16$0.20$3.00

Why 14.8 and not 15

The provider clips audio after 14.8 seconds, so a 15-second file would lose its last 0.2 seconds of speech. Sume refuses to send it. Measure the file you will upload and pass that exact number as duration_seconds, rather than an upper bound.

  • 14.8 and 14.79 both bill as 15 seconds.
  • 14.81 is refused with a validation error, and no reservation is made.
  • 5.0 bills as 5 seconds; 5.01 bills as 6.

A guard before you submit

This helper computes the billed amount and refuses lengths outside the window, so a bad value fails on your side first. The arithmetic uses Decimal to avoid float noise.

import math
from decimal import Decimal

LIST = {'480p': Decimal('0.05'), '768p': Decimal('0.08'), '1080p': Decimal('0.16')}
MARGIN = Decimal('1.25')

def lipsync_cost(seconds, resolution='768p'):
    if not 5 <= seconds <= 14.8:
        raise ValueError('audio must be 5 to 14.8 seconds')
    return LIST[resolution] * MARGIN * math.ceil(seconds)

print(lipsync_cost(14.8))            # 1.5000
print(lipsync_cost(14.8, '1080p'))   # 3.0000

Spending it well

At 768p a full-window line costs $1.50, and the same line at 1080p costs $3.00, so the resolution choice is worth more than the 0.2-second rounding. Draft at 768p, ship at 1080p if the screen needs it. A line longer than 14.8 seconds has to be split in two before you generate the audio; do not trim the audio to fit, because the API never clamps or cuts it for you.

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume