Pika launches H3 Flex text-to-video model with audio stems and up to 4K output
Pika says its H3 Flex text-to-video model offers exportable audio stems, up to 4K resolution and faster-than-real-time generation, starting at one cent per second.
Pika has launched H3 Flex, a text-to-video model that the presenter Matthew Berman, writing for Forward Future, says debuted as the top open-weight model on Design Arena's text-to-video board. The headline claims are exportable audio stems, resolutions up to 4K, and generation faster than real time. Those are the company's claims as relayed in the video.
Berman says the model is trained on MiniMax H3, the same video model that other developers have been building on. Pika's version is, in his words, faster and better overall. He cites a 15-second clip with a soundtrack produced in as little as 7 seconds.
Pricing starts at one cent per second. Berman also said the model accepts several inputs at once. In his demonstration he supplied five separate images and a WAV file of a thunderstorm, and the model combined them into one video with camera movement. He also said the output can be edited afterward, with changes made to one part of the video.
Berman said he has been running the model locally and that a high-end graphics card would probably be enough.
The examples Berman showed were two players playing ping-pong in a pool, one of them a frog, with lots of zooms and pans. He also showed an intro video with lots of camera movement and realism, where the hair looks beautiful. He described the model as very good at editor control.
What changed
Pika launched H3 Flex, a text-to-video model with exportable audio stems, output up to 4K and pricing starting at one cent per second.
Type: New capability · Applies to: H3 Flex
Why it matters
Lower per-second video pricing and audio stems could make generated video practical for production pipelines such as ads and social content.
What you can do
Teams producing video content could compare cost per second and editing controls against their current video vendor.
Original source
We have no link to the original source for this story yet.
Where we saw it
Forward Future (Matthew Berman)
AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works