Vendors
18 vendors and 131 models - one key, below official prices.
GPT-5 for reasoning and code, plus image generation.
Gemini for text, Nano Banana for images and Veo 3.1 for video.
Qwen-Image and Wan: images, plus 1080p video with audio.
Claude: strong coding and writing, large context.
Seedream for images and Seedance for video with consistent multi-shot scenes.
Grok Imagine: image generation and editing.
Kling: cinematic video with native audio and multi-shot scenes up to 15 seconds.
FLUX: photorealistic image generation with accurate text.
Hailuo and H3: image-to-video with lifelike motion and facial expression.
DeepSeek V4: code and reasoning with a 1M context at a low price per token.
Kimi K3: a flagship for long coding tasks, always reasons, 1M context.
Midjourney V7 and V8: the signature look, --ar and --sref flags in the prompt, edits from a reference.
Ideogram 3.0: legible text in images, posters and logos; a character from one portrait.
Recraft: 2x upscaling and background removal in one request.
Vidu Q3: up to 1080p, 16-second video from text or a frame, with sound.
PixVerse: fast short-form video from 360p to 1080p.
Runway Gen-4 Turbo: fast 5-second 720p video from text or a frame.
Jev: makes decisions instead of writing text - a choice, a score, yes/no with probabilities.