⭐ Football Update: Match Update & Schedules

Nano Banana 2.1

Krea

After Effects

FLUX 3 Image

Ideogram 4.5

GPT Image 2.5

Seedance 2.5

MiniMax H3

Introducing Video 2.0

Introducing Krea 3

Get Krea into ChatGPT

Krea Agent

Turn an idea into images and videos. Agent plans, creates, and refines them with you.

Open Agent

Create images

Nano Banana, GPT & more.

Create with plugins

Use Krea in your creative software.

Create videos

Seedance, MiniMax & more.

Make ads

Complete ad campaigns in seconds.

Connect your agent

Krea in Claude & ChatGPT.

Render as you draw

Create on a live canvas.

Create with Krea Agent

Explainer Video

Turn a topic into a narrated explainer for your faceless channel, with a paper-collage style.

Use template

Move the Location

Give an existing clip a new setting for your next ad or story, without filming again.

Use template

One Park, Four Seasons

The same cinematic composition reimagined from spring through winter

Use template

New Face, New City

Adapt one drink ad for different audiences with a new creator and location.

Use template

Indoors to Outdoors

Take your product demo outdoors and get a fresh version from the footage you already have.

Use template

Company Merch

Imagine ten different products with your own branding

Use template

Street Style Moment

Create street-style content for fashion Reels, with a new look for each post.

Use template

Home Walkthrough

Turn a property photo or render into a video tour for listings and client presentations.

Use template

Brand World Teaser

Introduce a product launch with a playful 3D teaser in your brand’s visual style.

Use template

Night Drift

Make cinematic drift clips for a car channel or an automotive campaign.

Use template

A new era

Introducing the intelligent Krea Agent. The future of generative media.

Try video models

Featured New

Seedance 2.5

ByteDance's flagship video model: cinematic 30-second clips at up to 1080p with 20% better prompt adherence, native synchronized audio, tagged image/video/audio references, region-level editing, and frame animation.

~850

Generate video

Featured New

MiniMax H3

Minimax frontier video model rendering at 2K. Generate from text, animate between start and end frames, or condition on tagged image, video, and audio references.

~500

Generate video

New

Gemini Omni Flash

Google's video-first multimodal model. Generate from text, a starting image, or tagged image references — with native speech and sound effects, plus single-clip video editing.

~750

Generate video

Featured New

Seedance 2.0

ByteDance frontier video model with cinematic motion, native synchronized audio, tagged image/video/audio references, and support for animating between start and end frames.

~1100

Generate video

New

Kling 3.0

Latest frontier model from Kling with native audio and extended durations up to 15 seconds.

~300

Generate video

New

Veo 3.1 Lite

Faster and more affordable version of the powerful Veo 3.1 model.

~200

Generate video

Featured New

Flux 3 Video

One model across image, video and audio. FLUX 3 renders up to 20 seconds in a single pass with multilingual speech, effects and ambience generated alongside the frames, and holds multiple scenes and camera angles together across cuts.

~650

Generate video

New

MiniMax H3 Max

MiniMax H3 Max is a post-trained variant of H3 tuned for stronger prompt adherence and better aesthetics, served through fal. Text-to-video, image-to-video with optional end frame, or reference-to-video with tagged image, video, and audio references — up to 15 seconds at 480p, 768p, or native 1080p.

~200

Generate video

New

MiniMax H3 Max Turbo

MiniMax H3 Max Turbo is fal’s speed- and cost-optimized H3 Max variant. Text-to-video, or image-to-video with an optional end frame — up to 15 seconds at 480p, 768p, or native 1080p. No reference-to-video on this tier.

~150

Generate video

New

LTX-2.5 Pro

Lightricks LTX-2.5 in its quality-optimized mode, for final output. Generates synchronized native audio in the same pass, with start and end frame support and optional camera motion.

~750

Generate video

New

LTX-2.5 Fast

Lightricks LTX-2.5 in its speed-optimized mode, built for rapid iteration. Generates synchronized native audio at every resolution, up to 20 seconds and 4K, with start and end frame support and optional camera motion.

~400

Generate video

New

LTX-2.3 22B

High-quality Lightricks video generation model with native audio and start and end frame support.

~150

Generate video

New

Kling 3.0 Turbo

Kling 3.0 Turbo text and image-to-video model with 720p/1080p output and durations up to 15 seconds.

~400

Generate video

New

Runway Gen-4.5

Latest frontier model from Runway with native text-to-video.

~450

Generate video

New

Seedance 2.0 Fast

Faster ByteDance Seedance 2 variant with cinematic motion, native synchronized audio, tagged image/video/audio references, and support for animating between start and end frames.

~900

Generate video

New

Seedance 2.0 Mini

ByteDance Seedance 2.0 Mini model with native audio, tagged image/video/audio references, and support for animating between start and end frames.

~550

Generate video

New

Gemini Omni Flash 1.1

Google's latest Gemini Omni video model. Generate from text, a starting image, or tagged image and short video references (up to three 3s clips) — with native speech and sound effects, plus single-clip video editing.

~750

Generate video

New

Grok Imagine 1.5

Grok Imagine 1.5 creates clips with synchronized audio, music, and sound effects. Text prompts generate a K2 start frame before animation.

~600

Generate video

New

HappyHorse

HappyHorse generates synchronized audio-video clips from text, image references, or a source video edit instruction.

~500

Generate video

New

FLUX 3 Max

Fal’s post-trained FLUX 3 variant improves prompt adherence and visual quality with lower latency. Generate up to 20 seconds at 720p or 1080p with synchronized audio, optionally guided by start and end images.

~650

Generate video

New

Flux Video Edit

Black Forest Labs' video editing model. Attach a clip up to 15 seconds and describe the change — remove, add or replace objects, restyle, change text, dialogue or setting. Length, framing and timing follow the source.

~100

Generate video

New

Kling o3

Advanced reasoning video model (720p). Supports image, element, and video references for precise creative control.

~300

Generate video

New

Kling o3 Pro

Advanced reasoning video model (1080p). Supports image, element, and video references for precise creative control.

~400

Generate video

New

Wan 3.0

Alibaba's latest generation video model: clips up to 30 seconds at up to 1080p with native synchronized audio, start and end frames, and up to 10 reference images, 5 reference videos, and 5 reference audios.

~550

Generate video

New

Wan 3.0 Prime

Faster version of Alibaba's Wan 3.0 video model: clips up to 30 seconds at up to 1080p with native synchronized audio, start and end frames, and up to 10 reference images, 5 reference videos, and 5 reference audios.

~500

Generate video

New

LTX-2 19B

Affordable medium-quality audio-video model from Lightricks with native sound generation.

~35

Generate video

New

Vidu Q3

New model excelling at anime.

~550

Generate video

Hailuo 2.3

Frontier model with dynamic motion.

~200

Generate video

Free

Hailuo 2.3 Fast

Cheapest medium-quality model. Best for most use cases.

~150

Generate video

Grok Imagine

Fast, high-quality video generation by xAI.

~300

Generate video

Kling o1

Intelligent video model that thinks before generating. Supports image, element, and video references for precise creative control.

~400

Generate video

Wan 2.6

Medium-quality model from Alibaba with improved quality and multi-shot support.

~350

Generate video

Seedance 1.0 Pro Fast

Fast and cheap model. Up to 12 seconds.

~85

Generate video

Veo 3.1

Best video model. Highest-quality frontier model with audio and reference images.

~1200

Generate video

Veo 3.1 Fast

Faster and more affordable version of the powerful Veo 3.1 model with audio.

~450

Generate video

Wan 2.5

Latest medium quality model from Alibaba.

~350

Generate video

Kling 2.6

Frontier model from Kling with native audio. Highest quality at a moderate price point.

~250

Generate video

Kling 2.5

Next-gen model with improved dynamics and enhanced style adaptation from Kling.

~250

Generate video

Kling 2.5 Turbo

Top-tier text-to-video generation with unparalleled motion fluidity and cinematic visuals.

~150

Generate video

Vidu Q2

High-quality model with reference images support.

~200

Generate video

Hailuo 02

Frontier model with dynamic motion.

~200

Generate video

Seedance 1.0 Pro

Fast, high-quality model from ByteDance.

~200

Generate video

Free

Wan 2.1

Fastest low-quality model with video Lora support.

~250

Generate video

Free

Wan 2.2

Fast, lower-quality model from Alibaba.

~200

Generate video

Runway Gen-4 Video

Medium quality model with a focus on cinematic visuals but weaker structure and motion.

~200

Generate video

Runway Gen-3

Old generation cinematic model with higher consistency.

~200

Generate video

Hailuo

High-quality model with camera control.

~300

Generate video

MiniMax 01-Live

High-quality model for animating people.

~300

Generate video

Ray 2

Older medium model with natural motion from Luma Labs.

~300

Generate video

Kling 1.0 Pro

Slow model with high control and 10s duration.

Generate video

Explore image models

Featured

Krea 2 Large

More powerful version of Krea 2 optimized for expressive photorealism.

Generate image

Featured

Krea 2 Medium

A smaller variant of Krea 2. Works best with illustrations and graphic design

Generate image

Featured

Krea 2 Turbo

The fastest Krea 2 model. Best for quickly iterating on expressive illustrations.

Generate image

Featured New

ChatGPT 2

Smartest model. Excellent prompt following. Best for complex diagrams, characters, and image editing.

~75

Generate image

Featured

Nano Banana Pro

Smartest model. World's best prompt adherence. Best model for complex tasks and image editing.

~100

Generate image

Featured New

Nano Banana 2

Google's latest flash image model (also known as Gemini 3.1 Flash Image) optimized for fast generation with support for up to 4K resolution.

~45

Generate image

Featured New

Nano Banana 2 Lite

Fast access to Google Nano Banana 2 Lite for text-to-image generation.

Generate image

Featured New

Luma UNI-1

Luma UNI-1 reasons through prompts before generating and supports optional visual references.

~30

Generate image

Featured New

Ideogram 4.0

Open image model focused on design, text rendering, and 2K photorealistic output.

~45

Generate image

New

Krea 3

Generate and edit images with image references, weighted styles, and precise edits.

~35

Generate image

New

GPT Image 2.6 Flare

GPT Image 2.6 Flare generates images with transparent background support, following the gpt-image-2 request shape.

~75

Generate image

New

Nano Banana 2.1

Nano Banana 2.1 delivers Pro-level quality at Flash speed, with improved product recontextualization, mask-based editing, and factuality over Nano Banana 2 and Pro.

Generate image

New

Seedream 5 Lite

Latest high-quality model from ByteDance with deep thinking, web search capabilities, and up to 14 style images.

~25

Generate image

New

Seedream 5 Pro

Flagship ByteDance model with precise local edits, multi-image fusion, strong text rendering, and up to 10 reference images.

~35

Generate image

New

Ideogram 4.5

Generate images and make edits with up to five image references.

~75

Generate image

New

Muse Image

Muse Image plans before it renders: it searches the web for real references, runs code for precise elements like charts and QR codes, and handles text-to-image, editing, and multi-reference composition (up to 10 images) in one model.

Generate image

New

Grok Imagine 2.0

Grok Imagine 2.0 generates images from text and edits up to 3 reference images. Choose a quality level and 1K or 2K output to trade detail against cost.

~45

Generate image

New

Recraft V4

Sharp, detailed images from Recraft with Standard and Pro modes.

~30

Generate image

New

Recraft V4.1 Flash

Recraft’s fast tier for raster art, illustration, and photography.

Generate image

New

Qwen Image 3

Alibaba’s third-generation image model. Reads prompts up to 4.5k tokens and renders text down to 10px across 12 languages, which suits posters, infographics, and storyboards. Attach up to 3 reference images to edit instead of generate.

~30

Generate image

New

Qwen Image 2.1

Generates and edits in one model, with native transparency for stickers, cutouts, and layered art. Keeps faces and products recognizable through outfit, material, and background changes, and composes from multiple references.

~30

Generate image

New

GPT Image 2.6 Sunburst

GPT Image 2.6 Sunburst is the faster GPT Image 2.6 variant. Five quality tiers trade latency against fidelity.

~75

Generate image

Free New

Z Image

Cheapest model. Medium quality photorealism at a budget. Realistic textures, weak diversity.

Generate image

Free New

Flux 2

FLUX.2 [dev] from Black Forest Labs. Enhanced realism and crisper text generation.

Generate image

New

Flux 3 Image

Early access to FLUX 3 Image. Combines up to ten references, supports grounded generation and native 4K output.

Generate image

Free

Krea 1

Most creative model with LoRAs. Excels at artistic photorealism. Realistic textures, weaker structure.

Generate image

GPT Image 2.5 Flare

GPT Image 2.5 Flare generates images with transparent background support, following the gpt-image-2 request shape.

~75

Generate image

Nano Banana

Smart model optimized for precise prompt adherence. Best for most tasks.

~30

Generate image

Flux 2 Klein

A lightweight 4 billion parameter version of Flux 2, optimized for speed while maintaining quality. Supports up to 2K resolution.

Generate image

Luma UNI-1 Max

Luma UNI-1 Max reasons more deeply than UNI-1 for higher-quality results and supports optional visual references.

~75

Generate image

Free

Qwen Image 2512

Latest Qwen model with enhanced human realism, finer natural detail for landscapes and textures, and improved text rendering with better layout and composition.