Oriel — AI & technology worth noticing, every day.

Pika-Labs/Pika-H3-Flex-Turbo-LoRA

AI model · https://huggingface.co/Pika-Labs/Pika-H3-Flex-Turbo-LoRA

First seen by Oriel on .

AI summary

Pika H3 Flex Turbo LoRA packages Pika's motion adapter with few-step MiniMax-H3 distillations, offering plain LoRAs for 2-, 4-, and 8-step video generation with synchronized audio.

  • Exact rank-concatenation preserves the stacked adapters mathematically, with reported per-layer relative error around 1–3e-7 and no custom nodes or model patches.
  • On H200 hardware, 4-step generation takes 40 seconds per five-second clip versus 60 seconds officially, while 2-step drafts take approximately 25 seconds.
  • The lineup includes Ref2VA and FL2VA variants, 3.9 GB standard files, and rank-32 Lite versions around 0.31 GB retaining 96.1% spectral energy.
  • ComfyUI usage requires a standard MiniMax-H3 workflow, LoraLoaderModelOnly at strength 1.0, matching step counts, CFG 1.0, and specified schedulers.
  • The 2-step file is draft-only because occasional scene failures and degraded audio are expected; final renders should use at least 4 steps when audio matters.

#lora #minimax-h3 #comfyui #video generation

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Likes

Total likes reported by the source.

  • Hugging Face Models: 174, latest observation ; +77 over observed period since

Downloads (30d)

Rolling 30-day downloads. A change in this value is not new downloads during the observed period.

  • Hugging Face Models: 90,541, latest observation ; +61,151 over observed period since

Aliases

SourceExternal IDURL
Hugging Face ModelsPika-Labs/Pika-H3-Flex-Turbo-LoRAhttps://huggingface.co/Pika-Labs/Pika-H3-Flex-Turbo-LoRA Read discussion