Technology

The models
behind the magic.

We pick the best AI for each task so you get one bill, one platform, and results that actually look and sound good.

Creative Generation — Creators Desk

Image Generation

  • Nano Banana 2

    Google DeepMind

    Premium quality photorealistic images. Flagship model.

  • Wan 2.7

    Alibaba

    Text-to-image and image editing. Strong on stylised output.

  • Qwen Image 2.0

    Alibaba / Qwen

    Value tier. Great for high-volume image generation.

  • Seedream v5.0 Lite

    ByteDance

    Fast generation. Good style range.

Video Generation

  • Veo 3.1

    Google DeepMind

    Premium tier. Cinematic quality, complex scene composition.

  • Sora 2

    OpenAI

    Premium tier. Temporal coherence, long clips.

  • Seedance 2.0

    ByteDance

    Premium tier. Fast render, expressive motion.

  • Other models

    Various

    Standard and Enhanced tiers — Wan video, Kling, and others.

Audio Intelligence

Audio Intelligence

  • Batch transcription — Speed tier

    Groq Whisper

    Small files up to 25 MB. Fastest processing, lowest cost.

  • Batch transcription — Heavy tier

    AssemblyAI

    Files up to 2.2 GB. Speaker labels, highest accuracy, custom vocabulary.

  • Live captioning

    AssemblyAI Real-time

    300ms latency streaming transcription for broadcasts and live events.

  • Text-to-Speech

    OpenAI TTS

    Multiple voices, standard and HD quality, instant generation.

  • Voice Q&A with AI

    OpenAI Realtime API

    After any transcript or voiceover: voice conversation with the AI. No typing, no lag.

Start building