Model providers

    Every model, by who built it.

    18 rosters covering 236 of the 296 AI models Versely runs. Each one is the complete list for that brand — every version, every tier, what it costs in credits, and the spec page that covers it.

    Kling39 models

    34 video models, 2 image models and 3 lipsync models

    eight product lines · 16 spec pages · 3–210 credits

    ByteDance31 models

    19 video models, 11 image models and 1 audio model

    eight product lines · 10 spec pages · 1–69 credits

    Wan27 models

    17 video models, 8 image models and 2 lipsync models

    five product lines · 14 spec pages · 2–50 credits

    Google25 models

    14 video models, 10 image models and 1 audio model

    five product lines · 10 spec pages · 2–64 credits

    Flux15 models

    15 image models

    four product lines · 8 spec pages · 1–8 credits

    Qwen13 models

    10 image models and 3 audio models

    four product lines · 6 spec pages · 1–8 credits

    LTX12 models

    10 video models and 2 lipsync models

    three product lines · 10 spec pages · 4–20 credits

    Recraft12 models

    12 image models

    three product lines · 3 spec pages · 4–30 credits

    OpenAI9 models

    5 video models and 4 image models

    three product lines · 3 spec pages · 2–50 credits

    PixVerse9 models

    9 video models

    five product lines · 7 spec pages · 9–75 credits

    VEED9 models

    3 video models and 6 lipsync models

    four product lines · 4 spec pages · 2–50 credits

    Happy Horse7 models

    7 video models

    two product lines · 4 spec pages · 18–28 credits

    Grok6 models

    3 video models, 2 image models and 1 audio model

    three product lines · 4 spec pages · 2–8 credits

    Runway6 models

    3 video models, 2 image models and 1 video-upscaling model

    three product lines · 4 spec pages · 3–50 credits

    Black Forest Labs4 models

    4 video models

    three product lines · 3 spec pages · 9–21 credits

    Hailuo4 models

    4 video models

    two product lines · 3 spec pages · 8–49 credits

    HiDream4 models

    4 image models

    two product lines · 3 spec pages · 1 credits

    Inworld4 models

    4 audio models

    two product lines · 3 spec pages · 3–5 credits

    Why only 18 rosters

    Versely runs models from around 50 providers, and most of them ship one or two. A roster needs enough in it to be worth reading as a list: a hub over three models is a shorter, worse version of the three model pages underneath it. So the bar here is six models in the catalog and at least three spec pages to link, and everything below it is covered on the model index instead — all 148 spec pages, grouped by what they make.

    A few names in the catalog are labels rather than brands, and the rosters here are grouped by who actually builds the model. Wan and Qwen are both Alibaba, as is the Happy Horse line; Kling is Kuaishou; LTX is Lightricks; Grok is xAI; Hunyuan is Tencent; the Flux image models and the FLUX.3 video models are both Black Forest Labs, filed under two different keys. Where that split would mislead, the roster says so rather than merging two product families that answer different questions.

    One subscription, every provider

    The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.