AI inference platform fal has introduced H3 Max Turbo, a faster and cheaper version of H3 Max just one week after the original model’s release. The new version remains based on MiniMax H3 post-training and is optimized for fal’s own inference stack. In the company’s official test for a 5-second, 768p video, generation latency fell from 3.49 seconds on H3 Max to 1.40 seconds on H3 Max Turbo, a roughly 2.5x speed gain. The tradeoff is a slight dip in quality. In fal’s internal benchmark, the original H3 Max was set at 100% quality, while Turbo scored 97%.
fal said prompt understanding, aesthetics, and overall output quality were largely preserved, and added that Turbo still ranked above the original MiniMax H3 and other video models included in the same test. Pricing also moved lower. The standard 768p rate for H3 Max Turbo is $0.04 per second, half of H3 Max’s regular $0.08 per second. fal is also offering a launch promotion with a 75% discount, bringing the cost down to $0.01 per second. At that rate, generating a one-minute video costs $0.6.
AI inference platform fal has launched H3 Max Turbo, a faster and cheaper version of H3 Max, just one week after the original release. The model remains based on MiniMax H3 post-training and has been optimized for fal’s own inference system.
Latency drops to 1.40 seconds
In fal’s official test for a 5-second, 768p video, generation latency fell from 3.49 seconds on H3 Max to 1.40 seconds on H3 Max Turbo, or about 2.5x faster.
The speed gain came with a small decline in quality. In fal’s internal test, the original H3 Max was rated at 100% quality, while Turbo scored 97%. The company said prompt understanding, aesthetics, and overall quality were largely preserved, and said Turbo still performed above the original MiniMax H3 and other video models included in the comparison.
Standard price cut in half, launch promotion lowers it to $0.01 per second
H3 Max itself was released on Aug. 27. One week later, fal moved again on both speed and pricing. The standard 768p price for Turbo is $0.04 per second, half of H3 Max’s regular $0.08 per second.
fal is also offering a 75% launch discount, bringing the rate down to $0.01 per second. At that price, generating a one-minute video costs $0.6.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
