Pika Labs publishes Pika H3 Flex, trending on Hugging Face
First merged summary.
Pika H3 Flex, a MiniMax-H3 post-train, is trending on Hugging Face with local 480p/768p video and stereo audio.
Pika Labs published Pika H3 Flex, a post-trained model built on MiniMax-H3, trending at number 29 on Hugging Face as of October 29, 2026. It generates video with synchronized stereo audio in six sampling steps and accepts up to 12 reference files through Ref2VA. Local weights cover 480p and 768p at 24 fps for requested lengths of 5 to 15 seconds. The Pika API adds 1080p, 2K and 4K plus audio stems. FL2VA covers text-to-video and first- or last-frame conditioning.
- Trending number 29 on Hugging Face.
- Local output is 480p or 768p at 24 fps, 5–15 seconds.
- Hosted API adds 1080p, 2K, 4K and audio stems.
- Ref2VA accepts up to 9 images, 3 videos and 3 audio clips.
- A 5-second request yields 124 frames, about 5.17 seconds, as 32 kHz stereo AAC MP4.
Timeline
dated events in the reporting · oldest first- Oct 29, 2026Pika Labs published Pika H3 Flex on Hugging Face. · Hugging Face trending models
What each outlet adds
beyond the shared facts- Hugging Face trending modelsLocal resolutions, API extras, Ref2VA limits and frame count.
- Products
- Pika API
- Organizations
- Hugging Face
Coverage
oldest first · one row per article · UTCOpen-H3-Regenerate-2K upscales MiniMax-H3 drafts to 2560×1440
Authors report a five-second clip falling from 87.26 seconds to 22.45 seconds on one B200 with the fast decoder.
Pika releases Pika H3 Flex open weights for video and audio
Local inference outputs native 480p or 768p with stereo audio; the hosted API adds up to 4K.