Video models
Wan 3.0 NEW
First on a public text-to-video leaderboard
21 credits / secondKling 4.0 NEW
Clips from 3 to 30 seconds in one pass
30 credits / secondGenjutsu NEW
Keeps the motion, camera and timing of your clip
205 credits / secondVeo 3.1
Native audio with dialogue
95 credits / secondVeo 3
Previous Veo generation
51 credits / secondSeedance 2.5
Native 30-second generation in one pass
27 credits / secondSeedance 2.0
Leads the quality rankings
26 credits / secondGrok Imagine
Clips from 4 to 12 seconds
11 credits / secondMiniMax H3
Native 2K at 24fps with stereo sound in one pass
17 credits / secondKling 3.0
Physical motion at 1080p
27 credits / secondKling 3.0 Motion Control
Copies body movement and gestures from a reference video
27 credits / secondImage models
GPT Image 2.5 NEW
Sharper detail and more natural light
19 credits / imageGrok Imagine Image 2 NEW
2K images at a low rate
8 credits / imageGPT Image 2 (medium)
Typography that stays readable at small sizes
14 credits / imageSeedream 5.0 Pro
Full layouts in 15 languages
14 credits / imageFLUX 2 [max]
Photorealistic default
16 credits / imageRecraft V3
Strong on long in-image text
6 credits / imageNano Banana
Fast edits
12 credits / imageNano Banana Pro
Photoreal skin and texture
38 credits / imageReal-ESRGAN
Classic upscaler
10 credits / imageNano Banana 2
Fast photoreal output
26 credits / imageFLUX schnell
Lowest image rate here
1 credit / imageAudio and voice
Text models
Gemini 2.5 Pro
A very long context
16 credits / 1k wordsGPT-5.5
The best brief adherence
47 credits / 1k wordsClaude Sonnet 5
The best style in longer forms
16 credits / 1k wordsGLM-5
Works well bilingually
6 credits / 1k wordsKimi K2.5
A long context
6 credits / 1k wordsGemini 3 Flash
Fast and cheap
5 credits / 1k wordsQwen3 Coder
Technical writing and documentation
8 credits / 1k wordsGPT-OSS 120B
Open weights
1 credit / 1k wordsDeepSeek V4
Strong reasoning
2 credits / 1k wordsClaude Haiku 4.5
Fast, natural sentences
8 credits / 1k words