Skip to content

Category

Video

Video generation and understanding. The fastest-moving frontier: seconds of coherent motion from a sentence. Reach for video when motion carries the message. Cost is counted rather than per , so iterate on short clips and spend the budget on the final render.

5 models in Video from 5 providers

R/video-sec Video is billed by the second of footage, so the cards below show a per-second price and no context window.

Top models

Leading video models

View all →
  • Hailuo Video 02 by MiniMax
    MiniMaxOn request

    Hailuo Video 02

    minimax/hailuo-video-02

    Cinematic short-form video generation with notably natural human motion.

    VideoStreaming

    — ctx · Price on request

  • HunyuanVideo by Tencent Hunyuan
    Tencent HunyuanOn request

    HunyuanVideo

    tencent/hunyuan-video

    Open-weight video generation: the reference open model for text-to-video research and production.

    VideoOpen weights

    — ctx · Price on request

  • Seedance 1.0 by ByteDance Doubao
    ByteDance DoubaoOn request

    Seedance 1.0

    bytedance/seedance-1.0

    Best-in-class motion coherence for short video, TikTok-era creative DNA in a generative model.

    Video

    — ctx · Price on request

  • Sora 2 by OpenAI
    OpenAIOn request

    Sora 2

    openai/sora-2

    Cinematic video generation with physics-aware scenes and audio-synchronised output.

    Video

    — ctx · Price on request

Browse all video models →

5 models in the directory

How to choose

Choosing a model in Video

  1. 01

    Cost scales per second

    Video is priced per generated second: short clips for iteration, longer renders only when the shot is right.

  2. 02

    Motion vs fidelity

    Some models prioritise temporal coherence, others frame quality. Preview both on your subject matter.

  3. 03

    Plan for review cycles

    Generation is not deterministic; budget several attempts per final shot.