Guides
Learn every tool, from your first image to your best one.
Video Tutorials
Placeholder links — real tutorial videos are on the way.
Visual Guide, Tool by Tool
Image
- 1.Search models and compare speed, quality, price, and reference-image support.
- 2.Describe the subject, style, lighting, and camera angle — the more specific, the better.
- 3.Set aspect ratio and resolution, then generate. The result is saved to your Gallery automatically.
Video
- 1.Choose a video model and describe the scene, motion, and mood you want.
- 2.Pick a duration and aspect ratio suited to where the clip will be used.
- 3.Generate and preview the clip, then download or regenerate with adjustments.
Speech
- 1.Type or paste the text you want spoken.
- 2.Choose a voice and language that fits your content.
- 3.Generate and download the audio file, ready to use.
Transcription
- 1.Upload an audio file you want converted to text.
- 2.Choose the source language for the most accurate result.
- 3.Generate and copy or download the transcript.
Meet the AI Models
Every model is good at something different — here is what each is best for.
Fast
Wan 2.7
Alibaba
Alibaba's image model. Create from text with an optional color palette, or edit with up to 9 reference images.
Fast create & edit
Seedream 5.0 Lite
ByteDance
ByteDance's next-gen lite model. Web-connected retrieval and strong complex-prompt comprehension.
Fast & capable
GPT Image 1 Mini
OpenAI
Efficient OpenAI image model. Great text rendering and editing with lower latency.
Fast & affordable
Nano Banana 2
Google's newest fast image model (Gemini 3.1 Flash Image). Quick generation with strong editing.
Fast editing
Nano Banana 2 Lite
Google's lightest, cheapest image model (Gemini 3.1 Flash Lite Image). Great for quick drafts.
Cheapest & fastest
Nano Banana
Fast generation with contextual understanding and editing capabilities (Gemini 2.5 Flash Image).
Fast generation
Riverflow V2 Fast
Sourceful
Speed-optimized reasoning model. Great for production deployments and quick iterations.
Production use
FLUX.2 Klein 4B
Black Forest Labs
Lightweight FLUX variant. Good quality at lower computational cost.
Efficient generation
Krea 2 Medium Turbo
Krea
Speed-optimized Krea model for rapid iteration.
Rapid iteration
GPT Image 1 Mini
OpenAI
Efficient dedicated OpenAI image model with lower latency.
Fast & affordable
Riverflow V2.5 Fast
Sourceful
Speed-optimized next-gen reasoning model from Sourceful.
Production use
Nano Banana 2 (Preview)
Preview channel for Nano Banana 2 (Gemini 3.1 Flash Image) — same model, early-access track.
Fast editing
Muse Image
Meta
Meta's image generation model, using its provider-selected output settings.
Fast image generation
Seedream 5.0 Flash
ByteDance
ByteDance's fastest, cheapest Seedream 5.0 tier. Built for high-volume and interactive use.
Fastest & cheapest
MAI-Image 2.6 Flash
Microsoft
Microsoft's lower-latency MAI-Image 2.6 tier. Comparable quality at higher throughput.
Fast & affordable
Recraft V4.1 Flash
Recraft
Recraft's speed and cost tier of the V4.1 family. Generates ~1K images in about 1.5 seconds.
Fast design
Standard
Luma Uni 1
Luma
Luma's image model. Create from text or edit one image with an instruction, in seven aspect ratios.
Create & edit
Seedream 4.5
ByteDance
ByteDance's latest. Strong portrait refinement, text rendering, and artistic expression.
Portraits & art
GPT Image 1
OpenAI
OpenAI's full-size image model. Strong prompt following with richer detail than Mini.
Balanced quality
Nano Banana Pro
Google's flagship image generation (Gemini 3 Pro Image). Advanced scene understanding.
Complex scenes
Recraft V4.1 Vector
Recraft
Specialized vector generation with clean lines, scalable designs, and SVG output.
Vector graphics
Grok Imagine
xAI
xAI's high-fidelity model. Photorealistic outputs with accurate textures and brands.
Photorealism
FLUX.2 Pro
Black Forest Labs
Black Forest Labs flagship. Sharp textures, stable lighting, consistent characters.
Consistency
FLUX.2 Flex
Black Forest Labs
Flexible FLUX variant balancing quality and speed for various use cases.
Versatile use
MAI-Image 2.5
Microsoft
Microsoft's image generation model. Balanced quality and speed.
General purpose
Krea 2 Medium
Krea
Balanced Krea model for everyday generation.
Balanced quality
Qwen Image 3
Qwen
Qwen's unified generation and editing model. Precise text rendering down to 10px.
Precise text & detail
GPT Image 1
OpenAI
OpenAI's dedicated image model. Strong prompt following with rich detail.
Balanced quality
Recraft V4 Vector
Recraft
Previous-gen vector generation with clean lines and scalable designs.
Vector graphics
Nano Banana Pro (Preview)
Preview channel for Nano Banana Pro (Gemini 3 Pro Image) — same model, early-access track.
Complex scenes
Recraft V4 Styles
Recraft
Reference-driven Recraft model for creating and applying visual styles.
Style references
Recraft V4 Styles Vector
Recraft
Reference-driven vector style generation with SVG output.
Styled vector art
Ming Image Design Layer
inclusionAI
inclusionAI's layer-decomposition model — splits a flattened design into separate RGBA layers.
Layer decomposition
MAI-Image 2.6
Microsoft
Microsoft's precision-tier image model. Design-ready visuals with improved reference support.
General purpose
Premium
Luma Uni 1 Max
Luma
Luma's higher-quality tier. High-fidelity text-to-image and single-image edits.
Highest fidelity
Wan 2.7 Pro
Alibaba
Alibaba's higher-fidelity tier. Richer detail and cleaner composition, with 4K text-to-image.
4K detail & editing
Seedream 5.0 Pro
ByteDance
ByteDance's flagship. Create from text or edit with up to 10 reference images — precise edits, lifelike scenes, sharp text.
Create & edit
Seedream 5.0 Pro
ByteDance
ByteDance's flagship editing model. Precise editing control, lifelike scenes, natural rendering.
Precise editing
GPT Image 2
OpenAI
OpenAI's most powerful image model. Exceptional prompt following and photorealism.
Best quality
GPT Image 2.5 Sunburst
OpenAI
OpenAI's newest image model. Same top-tier quality and pricing as GPT Image 2.
Best quality
Recraft V4.1 Pro
Recraft
Professional-grade design model with enhanced quality and precision for enterprise work.
Professional design
Recraft V4.1 Pro Vector
Recraft
Premium vector generation with professional-grade precision and detail.
Professional vector art
Grok Imagine 2.0
xAI
xAI's next-gen image and editing model. Sharper detail and more reliable reference editing.
Photorealism & editing
Riverflow V2 Pro
Sourceful
Multi-step reasoning image model. Top-tier control with 4K output support.
Max control
FLUX.2 Max
Black Forest Labs
Maximum resolution and quality FLUX model. Best for large format outputs.
Maximum quality
MAI-Image 2.5 Pro
Microsoft
Microsoft's flagship image model with enhanced detail and fidelity.
High fidelity
Krea 2 Large
Krea
Krea's largest model. Maximum detail and creative range.
Maximum detail
Qwen Image 3 Pro
Qwen
Qwen's flagship generation and editing model. Richer world knowledge and finer detail.
Precise text & detail
GPT Image 2
OpenAI
OpenAI's dedicated image generation model. Exceptional prompt following and photorealism.
Best quality
Riverflow V2.5 Pro
Sourceful
Sourceful's next-gen reasoning image model. Improved control over V2.
Max control
Recraft V4.1 Utility Pro
Recraft
Professional utility design model with enhanced precision.
Pro icons & UI
Recraft V4 Pro Vector
Recraft
Previous-gen premium vector generation with professional precision.
Professional vector art
Recraft V4 Pro
Recraft
Previous-gen professional design model with enhanced quality.
Professional design
Recraft V4 Styles Pro
Recraft
Higher-fidelity reference-driven style generation from Recraft.
Professional styles
Recraft V4 Styles Pro Vector
Recraft
Higher-fidelity reference-driven vector generation with SVG output.
Professional vector art
FLUX.3 Image
Black Forest Labs
Black Forest Labs' flagship image model. Text-to-image and multi-reference editing with up to 10 input images.
Flagship quality & editing
GPT Image 2.5 Flare
OpenAI
OpenAI's speed-oriented GPT Image 2.5 tier. Built for high-volume everyday generation.
Speed-oriented
Design & Vector
Recraft V4.1
Recraft
Design-first model. Strong prompt adherence, art-directed composition, text rendering.
Design & logos
Recraft V4.1 Utility
Recraft
Utility-focused design model. Icons, UI elements, and functional graphics.
Icons & UI
Recraft V4
Recraft
Previous-gen design-first model with strong prompt adherence.
Design & logos
Recraft V3
Recraft
Recraft's earlier design model. Reliable text rendering and composition.
Design & logos
Ming Image Design
inclusionAI
inclusionAI's graphic-design model, focused on legible text rendering inside generated images.
Graphic design & text