Model library
15 model series, 38 variants — all in the same workspace on the same credit balance. No vendor accounts, no separate subscriptions. Pick a series to see what it's good at and try it right on the page.
ByteDance's audio-video series — cinematic multi-shot clips from the 2.0 flagship to fast 1.0 tiers.
Google DeepMind's Veo 3.1 — cinematic video with natively generated audio.
Kuaishou's Kling — multi-shot storytelling with native multilingual audio.
Alibaba's Wan 2.7 — director-level control from generation to instruction-based editing.
MiniMax's Hailuo 2.3 Fast — animate a still with natural, physical motion.
PixVerse v6 — multi-shot clips with synchronized audio from a single prompt.
xAI's Grok Video 1.5 — quick, affordable clips with audio in the same pass.
Alibaba's HappyHorse 1.1 — cinematic motion with improved consistency and audio.
Google's Nano Banana series — Gemini image models from fast edits to Pro-grade accuracy.
OpenAI's GPT Image — production-grade generation with strong text rendering.
Midjourney on the web — five versions, no Discord required.
ByteDance's Seedream — dense text rendering and multi-image consistency.
xAI's Grok Imagine — expressive image generation from the Grok family.
Black Forest Labs' FLUX.1 — open-weight generation from instant drafts to reference editing.