Search the full model catalog.

GPT-6 Luna is OpenAI’s efficient GPT‑6 model for text tasks with image understanding. Use it for chat, coding help, and screenshot analysis at scale.

OpenAI GPT-6 Sol is a long-context reasoning model for coding and agent workflows. It accepts text with optional images and returns text responses.

Google’s Lyria 3.5 generates full-length songs from text prompts and optional images. It returns 44.1 kHz stereo MP3 audio plus lyrics text.

Claude Fable 5.1 is Anthropic's top-tier long-context model for agentic coding and deep research. It can read images or PDFs and replies in text.

Claude Opus 5.5 is a long-context reasoning model for agentic coding and knowledge work. It can read images and PDFs and returns structured, high-detail text answers.

Alibaba’s Qwen Image 3.0 Pro generates and edits PNG images from long prompts and up to 3 reference images, built for dense layouts and small text.

Alibaba's Qwen Image 3.0 generates or edits PNG images from long prompts and up to 3 reference images. It's built for dense layouts, crisp text, and 2K output.

Edit a photo with one instruction and up to 3 input images. This uncensored Qwen-based editor keeps identity and framing while changing objects or text.

Create uncensored images from a written scene prompt in 1K or 2K, with preset aspect ratios. Optional prompt expansion and a fixed seed help consistency.

Generate uncensored images from a single prompt at up to 1536px. Choose an aspect ratio, optionally expand your prompt, and set a seed for repeatability.

Storm Dance turns one portrait into a stylized crowd dance clip, keeping the person centered. Pick Black, Emerald, or Pink looks, then export 720p or 1080p.

Qwen-Image-2.1 is Qwen’s open-weight 7B model for text-to-image and image editing. It supports 2K PNG output, including native RGBA transparency.

Create and edit images with Alibaba’s Wan 2.7 Image Pro, with 4K text-to-image, multi-image reference editing, and coherent sequential image sets.

Alibaba Wan 2.7 Image generates or edits PNG images from text and up to 9 reference images. It supports 1K or 2K output and coherent image sets up to 12 images.

FastH3 V2 is an eight-forward distilled MiniMax H3 model that generates synchronized video and stereo audio from a text prompt. This H200-backed 8-step checkpoint supports native 5.167-14.375 second T2AV clips at 24 FPS; FL2VA and Ref2VA were not distilled.

Create 2–30s reference-guided videos from a prompt plus up to 10 images, 5 clips, and 5 audio tracks. Export 1080p, 2K, or 4K MP4.

Generate 5–30 second videos from a written shot prompt, with optional first and last frame keyframes, audio on/off, and 1080p to 4K output.

Generate 2–30s 480p–1080p MP4 videos from a prompt plus up to 10 reference images, with optional reference video and audio for stronger identity control.

Create 5–30s MP4 videos from a prompt, optionally anchored by first and last frame images with optional audio. Powered by Alibaba Wan 3.0 Prime.

Generate uncensored reference-to-video clips with Wan 3.0 Pro R2V. Add up to 10 images plus optional video and audio references, and export 1080p–4K MP4.