Pika H3 Flex video model trends on Hugging Face
First merged summary.
Pika Labs’ MiniMax-H3-based Pika H3 Flex is trending on Hugging Face, with local 480p or 768p video and stereo audio.
Pika Labs published Pika H3 Flex, a post-trained model built on MiniMax-H3, listed as trending number 29 on Hugging Face. It generates video with synchronized stereo audio in six sampling steps. FL2VA covers text-to-video and first- or last-frame conditioning, and Ref2VA accepts up to 12 reference files. 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.
- Built on MiniMax-H3 and sampled in six steps.
- Ref2VA accepts up to 9 images, 3 videos and 3 audio clips.
- Local output is 24 fps MP4 with 32 kHz stereo AAC.
- A 5-second request produces 124 frames, about 5.17 seconds.
- The hosted API adds 1080p, 2K, 4K and audio stems.
Timeline
dated events in the reporting · oldest first- Oct 29, 2026Pika H3 Flex listed trending number 29 · Hugging Face trending models
What each outlet adds
beyond the shared facts- Hugging Face trending modelsLocal resolutions, API up to 4K, and reference-file limits.
- Vendors
- Pika Labs
- Products
- Pika API
- Organizations
- Hugging Face
Coverage
oldest first · one row per article · UTCPRISM merge of Minimax-H3 Singularity and 10Eros-Max trends on Hugging Face
The fp16 file merges two community Minimax-H3 finetunes without mergekit, and the card adds later int8 tests.
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.