# Krea: English Model APIs

## Model APIs

### Getting Started

- [Krea API reference: models, endpoints, and schemas](https://www.krea.ai/docs/api-reference/introduction.md): Browse every Krea API model and endpoint with request formats, parameters, authentication details, and response examples for faster integration.
- [Import the Krea API into Postman](https://www.krea.ai/docs/api-reference/postman.md): Import the Krea OpenAPI spec into Postman to test API endpoints, configure bearer token authentication, and explore request schemas without writing code.

### General

- [List jobs](https://www.krea.ai/docs/api-reference/general/list-jobs.md): List jobs with pagination and filtering. Returns jobs belonging to the authenticated user.
- [Get a job by ID](https://www.krea.ai/docs/api-reference/general/get-a-job-by-id.md)
- [Delete a job by ID](https://www.krea.ai/docs/api-reference/general/delete-a-job-by-id.md)

### Krea

- [Krea 2 Medium](https://www.krea.ai/docs/api-reference/krea/krea-2-medium.md): Best for expressive illustrations.
- [Krea 2 Large](https://www.krea.ai/docs/api-reference/krea/krea-2-large.md): Best for expressive photorealism.
- [Krea 2 Turbo](https://www.krea.ai/docs/api-reference/krea/krea-2-turbo.md): Fastest medium quality Krea 2 model.

### Usage

- [List workspace usage](https://www.krea.ai/docs/api-reference/usage/list-workspace-usage.md): The workspace's completed metered jobs in a window, newest first, each with the compute units it used. It aggregates nothing: group and sum `compute_units` in your own pipeline.

### Assets

- [List assets](https://www.krea.ai/docs/api-reference/assets/list-assets.md): List assets with pagination.
- [Upload an asset](https://www.krea.ai/docs/api-reference/assets/upload-an-asset.md): Upload a new asset (image, video, audio, or 3D model). Accepts multipart/form-data with a file and optional description. Maximum file size: 75MB.
- [Get an asset](https://www.krea.ai/docs/api-reference/assets/get-an-asset.md): Get a single asset by ID
- [Delete an asset](https://www.krea.ai/docs/api-reference/assets/delete-an-asset.md): Delete an asset.

### Styles

- [Train a custom style (LoRA)](https://www.krea.ai/docs/api-reference/styles/train-a-custom-style-lora.md): Train a custom style (LoRA) with simplified parameters. Uses intelligent defaults based on the training type (Default/Style/Object/Character).
- [Search styles](https://www.krea.ai/docs/api-reference/styles/search-styles.md): List styles with pagination and filtering. Returns public styles, user-created styles, or shared styles based on the filter parameter.
- [Get a single style by ID](https://www.krea.ai/docs/api-reference/styles/get-a-single-style-by-id.md): Get a single style by ID. Returns public styles or your own private styles.
- [Update a style](https://www.krea.ai/docs/api-reference/styles/update-a-style.md): Update style properties. Currently supports updating the public visibility of a style you own.
- [Get a shareable link for a style](https://www.krea.ai/docs/api-reference/styles/get-a-shareable-link-for-a-style.md): Generate or retrieve a shareable link for a style you own. The link allows others to access the style.
- [Share a style with your workspace](https://www.krea.ai/docs/api-reference/styles/share-a-style-with-your-workspace.md): Share a style you own with your API workspace. All members of the workspace will be able to use the style.
- [Remove a style from your workspace](https://www.krea.ai/docs/api-reference/styles/remove-a-style-from-your-workspace.md): Remove a style you own from your API workspace. Workspace members will no longer have access to the style.

### Node Apps

- [List runnable node apps](https://www.krea.ai/docs/api-reference/node-apps/list-runnable-node-apps.md): List node apps visible to the authenticated user that can be executed through the public API.
- [List runnable node app versions](https://www.krea.ai/docs/api-reference/node-apps/list-runnable-node-app-versions.md): List versions of a node app visible to the authenticated user that can be executed through the public API.
- [Get a node app by version ID](https://www.krea.ai/docs/api-reference/node-apps/get-a-node-app-by-version-id.md): Get a single node app by version ID. Returns the node app details including input/output schemas.
- [Execute a node app](https://www.krea.ai/docs/api-reference/node-apps/execute-a-node-app.md): Execute a node app by version ID. The request body should contain the input data as a JSON object matching the node app's input schema. Returns a job that can be tracked using /jobs/{id}.

### Utilities

- [Export a 3D model](https://www.krea.ai/docs/api-reference/utilities/export-a-3d-model.md): Export a completed 3D generation job's model to OBJ, FBX, STL, or PLY. Pass the job id and target format. The exported model is returned as a ZIP file.

### Image

- [Ideogram 4.5](https://www.krea.ai/docs/api-reference/image/ideogram-45.md): Generate images or edit source images with prompts and up to five image references. Very low quality requires an image. Set preserve_source_size to keep the first source image dimensions; aspect_ratio and resolution are then ignored.
- [Flux.1 Dev](https://www.krea.ai/docs/api-reference/image/flux1-dev.md): Fast model. Best for LoRAs.
- [Flux Kontext Dev](https://www.krea.ai/docs/api-reference/image/flux-kontext-dev.md): Frontier model designed for image editing, optimized for Krea
- [Nano Banana Pro](https://www.krea.ai/docs/api-reference/image/nano-banana-pro.md): World's most intelligent model.
- [Nano Banana 2](https://www.krea.ai/docs/api-reference/image/nano-banana-2.md): Cheaper quality version of Nano Banana.
- [Nano Banana 2 Lite](https://www.krea.ai/docs/api-reference/image/nano-banana-2-lite.md): Fast Google image model.
- [Nano Banana](https://www.krea.ai/docs/api-reference/image/nano-banana.md): Most versatile intelligent model.
- [Flux 1.1 Pro](https://www.krea.ai/docs/api-reference/image/flux-11-pro.md): Advanced yet efficent model from BFL.
- [Flux 1.1 Pro Ultra](https://www.krea.ai/docs/api-reference/image/flux-11-pro-ultra.md): BFL's highest quality text to image model.
- [Ideogram 2.0A Turbo](https://www.krea.ai/docs/api-reference/image/ideogram-20a-turbo.md): Distilled version of Ideogram 2.0
- [Ideogram 3.0](https://www.krea.ai/docs/api-reference/image/ideogram-30.md): Highly aesthetic, general-purpose model.
- [Runway Gen-4 Image](https://www.krea.ai/docs/api-reference/image/runway-gen-4-image.md): Cinematic image model with references.
- [ChatGPT Image](https://www.krea.ai/docs/api-reference/image/chatgpt-image.md): Original ChatGPT image model.
- [ChatGPT 2](https://www.krea.ai/docs/api-reference/image/chatgpt-2.md): Smart powerful model for editing.
- [GPT Image 2.5 Flare](https://www.krea.ai/docs/api-reference/image/gpt-image-25-flare.md): GPT Image 2.5 with transparent background support.
- [GPT Image 2.5 Sunburst](https://www.krea.ai/docs/api-reference/image/gpt-image-25-sunburst.md): GPT Image 2.5, optimized for speed.
- [Grok Imagine 2.0](https://www.krea.ai/docs/api-reference/image/grok-imagine-20.md): xAI's image model with selectable quality and 2K output.
- [Grok Imagine 2.0](https://www.krea.ai/docs/api-reference/image/grok-imagine-20-1.md): xAI's image editing model with natural language edit instructions
- [Muse Image](https://www.krea.ai/docs/api-reference/image/muse-image.md): Meta’s agentic image model — it reasons about the prompt, pulls references from the web, and verifies the render before returning it.
- [Qwen Image 3](https://www.krea.ai/docs/api-reference/image/qwen-image-3.md): Dense layouts with small, legible text.
- [Qwen Image 2.1](https://www.krea.ai/docs/api-reference/image/qwen-image-21.md): Transparent assets and identity-safe edits.
- [Luma UNI-1](https://www.krea.ai/docs/api-reference/image/luma-uni-1.md): Reasoning image model for prompt-faithful text to image.
- [Recraft V4.1 Flash](https://www.krea.ai/docs/api-reference/image/recraft-v41-flash.md): Recraft’s fast tier for raster art, illustration, and photography.
- [Seedream 4](https://www.krea.ai/docs/api-reference/image/seedream-4.md): New high quality model for photorealism and text rendering
- [Seedream 5 Lite](https://www.krea.ai/docs/api-reference/image/seedream-5-lite.md): Medium quality model with reasoning and web search.
- [Seedream 5 Pro](https://www.krea.ai/docs/api-reference/image/seedream-5-pro.md): Highest quality model with precise editing control.
- [Qwen Image 2512](https://www.krea.ai/docs/api-reference/image/qwen-image-2512.md): Affordable medium quality model.
- [Z Image](https://www.krea.ai/docs/api-reference/image/z-image.md): Fastest model. Realistic, low diversity.
- [Ideogram 4.5 Precise](https://www.krea.ai/docs/api-reference/image/ideogram-45-precise.md): Precisely edit an image while preserving its resolution and surrounding details. Supports optional masks and reference images. Use generate_image with this model for both masked and unmasked edits.
- [SeedEdit](https://www.krea.ai/docs/api-reference/image/seededit.md): New frontier model designed for image editing

### Image Enhance

- [Krea Enhance](https://www.krea.ai/docs/api-reference/image-enhance/krea-enhance.md): Cheap creative enhancer for adding new details. Max 8K.
- [Krea Enhance Legacy](https://www.krea.ai/docs/api-reference/image-enhance/krea-enhance-legacy.md): Krea's generative legacy enhancer. Max 4K.
- [Topaz Image Upscale](https://www.krea.ai/docs/api-reference/image-enhance/topaz-image-upscale.md): Most powerful creative enhancer. Max 16K.
- [Topaz Standard](https://www.krea.ai/docs/api-reference/image-enhance/topaz-standard.md): Powerful faithful upscaler. Up to 22K resolution. Best for photos.
- [Topaz Bloom](https://www.krea.ai/docs/api-reference/image-enhance/topaz-bloom.md): Creative upscaler that adds new details. Max 10K.
- [Topaz Bloom 2](https://www.krea.ai/docs/api-reference/image-enhance/topaz-bloom-2.md): Creative upscaler for GenAI images, reimagined. Max 10K.
- [Topaz Wonder 3.5](https://www.krea.ai/docs/api-reference/image-enhance/topaz-wonder-35.md): Detail-faithful generative upscaler. Max 16K.
- [Magnific Creative](https://www.krea.ai/docs/api-reference/image-enhance/magnific-creative.md): Best creative enhancer. Max 10K.
- [Magnific Precise](https://www.krea.ai/docs/api-reference/image-enhance/magnific-precise.md): More faithful variant of Magnific. Max 10K.

### Video

- [Kling 2.5](https://www.krea.ai/docs/api-reference/video/kling-25.md): Improved dynamics and style adaptation.
- [Kling 2.6](https://www.krea.ai/docs/api-reference/video/kling-26.md): Frontier model with native audio.
- [Kling 3.0](https://www.krea.ai/docs/api-reference/video/kling-30.md): Frontier model with audio and up to 15s.
- [Kling o1](https://www.krea.ai/docs/api-reference/video/kling-o1.md): Intelligent model that thinks before generating.
- [Hailuo](https://www.krea.ai/docs/api-reference/video/hailuo.md): High-quality model with camera control.
- [Hailuo 02](https://www.krea.ai/docs/api-reference/video/hailuo-02.md): Frontier model with dynamic motion.
- [Hailuo 2.3](https://www.krea.ai/docs/api-reference/video/hailuo-23.md): Frontier model with dynamic motion.
- [Hailuo 2.3 Fast](https://www.krea.ai/docs/api-reference/video/hailuo-23-fast.md): Cheapest medium-quality model.
- [MiniMax H3](https://www.krea.ai/docs/api-reference/video/minimax-h3.md): Frontier 2K model with references and frame animation.
- [MiniMax H3 Max](https://www.krea.ai/docs/api-reference/video/minimax-h3-max.md): Post-trained H3 variant with stronger prompt adherence, better aesthetics, and image/video/audio references.
- [MiniMax H3 Max Turbo](https://www.krea.ai/docs/api-reference/video/minimax-h3-max-turbo.md): Faster, cheaper H3 Max tier with the same prompt adherence, at 480p, 768p, or native 1080p up to 15 seconds.
- [minimax_h3max_camera](https://www.krea.ai/docs/api-reference/video/minimax_h3max_camera.md)
- [Wan 2.1](https://www.krea.ai/docs/api-reference/video/wan-21.md): Fastest low-quality model with LoRA.
- [Wan 2.2](https://www.krea.ai/docs/api-reference/video/wan-22.md): Fast, lower-quality model from Alibaba.
- [Wan 2.5](https://www.krea.ai/docs/api-reference/video/wan-25.md): Latest medium quality model from Alibaba.
- [Wan 3.0](https://www.krea.ai/docs/api-reference/video/wan-30.md): Latest Wan with 30s clips, native audio, and image, video, and audio references.
- [Veo 2](https://www.krea.ai/docs/api-reference/video/veo-2.md): Expensive high-quality model from Google.
- [Veo 3](https://www.krea.ai/docs/api-reference/video/veo-3.md): Older version of Veo 3.1.
- [Veo 3 Fast](https://www.krea.ai/docs/api-reference/video/veo-3-fast.md): Faster, affordable Veo 3 with audio.
- [Veo 3.1](https://www.krea.ai/docs/api-reference/video/veo-31.md): Best video model. Highest quality with audio.
- [Veo 3.1 Fast](https://www.krea.ai/docs/api-reference/video/veo-31-fast.md): Faster, affordable Veo 3.1 with audio.
- [Veo 3.1 Lite](https://www.krea.ai/docs/api-reference/video/veo-31-lite.md): Faster, affordable Veo 3.1.
- [Gemini Omni Flash](https://www.krea.ai/docs/api-reference/video/gemini-omni-flash.md): Multimodal video with native audio.
- [Gemini Omni Flash 1.1](https://www.krea.ai/docs/api-reference/video/gemini-omni-flash-11.md): Multimodal video with native audio.
- [Seedance 1.0 Pro](https://www.krea.ai/docs/api-reference/video/seedance-10-pro.md): Fast, high-quality model from ByteDance.
- [Seedance 1.0 Pro Fast](https://www.krea.ai/docs/api-reference/video/seedance-10-pro-fast.md): Fast and cheap model. Up to 12 seconds.
- [Seedance 2.0*](https://www.krea.ai/docs/api-reference/video/seedance-20*.md): Flagship video model with native audio, references, and frame animation.
- [Seedance 2.0 Fast*](https://www.krea.ai/docs/api-reference/video/seedance-20-fast*.md): Faster Seedance 2 with native audio, references, and frame animation.
- [Seedance 2.0 Mini](https://www.krea.ai/docs/api-reference/video/seedance-20-mini.md): Cost-efficient Seedance 2 model for scalable video production.
- [Seedance 2.5*](https://www.krea.ai/docs/api-reference/video/seedance-25*.md): Most advanced Seedance with 30s clips, native audio, and up to 50 references. Default choice for most video generations — quality is head and shoulders above other models, it supports durations up to 30 seconds, and it outputs up to 1080p. Reach for this first unless the user asks for a specific mod…
- [LTX-2.3 22B](https://www.krea.ai/docs/api-reference/video/ltx-23-22b.md): High-quality audio-video model from Lightricks.
- [LTX-2.5 Pro](https://www.krea.ai/docs/api-reference/video/ltx-25-pro.md): Quality-optimized audio-video model from Lightricks.
- [LTX-2.5 Fast](https://www.krea.ai/docs/api-reference/video/ltx-25-fast.md): Speed-optimized Lightricks model with 20s clips up to 4K.
- [Vidu Q3](https://www.krea.ai/docs/api-reference/video/vidu-q3.md): New model excelling at anime.
- [Ray 2](https://www.krea.ai/docs/api-reference/video/ray-2.md): Older model with natural motion.
- [Runway Gen-4 Video](https://www.krea.ai/docs/api-reference/video/runway-gen-4-video.md): Cinematic visuals, weaker structure.
- [Runway Gen-4.5](https://www.krea.ai/docs/api-reference/video/runway-gen-45.md): Frontier model with native text-to-video.
- [Grok Imagine](https://www.krea.ai/docs/api-reference/video/grok-imagine.md): Fast, high-quality video generation by xAI.
- [Grok Imagine 1.5](https://www.krea.ai/docs/api-reference/video/grok-imagine-15.md): Text and image-to-video with native audio by xAI.
- [Flux 3 Video](https://www.krea.ai/docs/api-reference/video/flux-3-video.md): Up to 20s with synchronized sound.
- [Flux Video Edit](https://www.krea.ai/docs/api-reference/video/flux-video-edit.md): Edit any clip with a text prompt.

### Video Enhance

- [Topaz Video Upscale](https://www.krea.ai/docs/api-reference/video-enhance/topaz-video-upscale.md): Topaz's powerful AI upscaler. Add frames and grain, fix focus, and more. Max 8K.
- [Topaz Starlight Precise](https://www.krea.ai/docs/api-reference/video-enhance/topaz-starlight-precise.md): A diffusion-based AI upscaler from Topaz that transforms low-resolution and degraded video into HD quality. Max 4K.
- [Topaz Starlight Precise 2.5](https://www.krea.ai/docs/api-reference/video-enhance/topaz-starlight-precise-25.md): The latest diffusion-based AI upscaler from Topaz. Sharper results with fewer artifacts than the original Starlight. Max 4K.
- [Topaz Astra](https://www.krea.ai/docs/api-reference/video-enhance/topaz-astra.md): Creative video enhancer for upscaling AI-generated videos up to 4K resolution.
- [Topaz Astra 2](https://www.krea.ai/docs/api-reference/video-enhance/topaz-astra-2.md): Creative diffusion-based video upscaler with prompt guidance, creativity, and sharpness controls. Max 4K.
- [Topaz Hyperion 2.5](https://www.krea.ai/docs/api-reference/video-enhance/topaz-hyperion-25.md): Topaz's SDR to HDR conversion model. Expands SDR footage to HDR (BT.2100 PQ) with 10-bit output. Max 4K.
- [SeedVR 2](https://www.krea.ai/docs/api-reference/video-enhance/seedvr-2.md): Fast video upscaler. Max 4K.
- [Flux Video Upscale](https://www.krea.ai/docs/api-reference/video-enhance/flux-video-upscale.md): Super-resolution with precise or creative modes. Sources up to 2560px and 20s; output up to 3x and 4K.
- [Runway Ruby](https://www.krea.ai/docs/api-reference/video-enhance/runway-ruby.md): Runway's SDR→HDR converter. True HDR from any clip up to 30s — resolution, pixels, and audio untouched.

### 3D

- [Hunyuan3D 3.1 Pro](https://www.krea.ai/docs/api-reference/3d/hunyuan3d-31-pro.md): High‑fidelity text or image‑to‑3D with optional PBR materials, multi‑view inputs, and adjustable face count for cleaner meshes.
- [TRELLIS 2](https://www.krea.ai/docs/api-reference/3d/trellis-2.md): Native 3D generative model for high-quality 3D asset creation
- [Hunyuan3D-2.1](https://www.krea.ai/docs/api-reference/3d/hunyuan3d-21.md): High quality 3D model generator
- [Tripo](https://www.krea.ai/docs/api-reference/3d/tripo.md): High quality 3D model generator
- [Tripo H3.1](https://www.krea.ai/docs/api-reference/3d/tripo-h31.md): Tripo's flagship generator: sharp geometry from text or image at low cost.
- [Rodin V2.5](https://www.krea.ai/docs/api-reference/3d/rodin-v25.md): Detailed, well-structured meshes with PBR materials from text or image.
- [Rodin V2.5 Fast](https://www.krea.ai/docs/api-reference/3d/rodin-v25-fast.md): Rodin V2.5 tuned for speed and cost, for quick iteration on shapes.
- [Hunyuan3D 3.1 Rapid](https://www.krea.ai/docs/api-reference/3d/hunyuan3d-31-rapid.md): Fast, inexpensive tier of Hunyuan3D 3.1 for text or image-to-3D.
- [Meshy V7](https://www.krea.ai/docs/api-reference/3d/meshy-v7.md): Production-ready meshes from text or image, with clean topology and PBR-quality textures.
- [TRELLIS](https://www.krea.ai/docs/api-reference/3d/trellis.md): High quality 3D model generator

### Audio

- [Seed Audio 1.0](https://www.krea.ai/docs/api-reference/audio/seed-audio-10.md): Generate speech, dialogue, and sound design from a text prompt with Seed Audio 1.0.
- [ElevenLabs TTS](https://www.krea.ai/docs/api-reference/audio/elevenlabs-tts.md): Convert text to natural-sounding speech with ElevenLabs voices. Defaults to eleven_multilingual_v2; pass model_id: eleven_v4 for ElevenLabs' most expressive model (best voice cloning, 90+ languages) or eleven_v4_turbo for low-latency delivery.
- [ElevenLabs Music](https://www.krea.ai/docs/api-reference/audio/elevenlabs-music.md): Generate a song or instrumental track from a text prompt with ElevenLabs Music. Vocals are included by default; set force_instrumental for instrumental-only output.
- [ElevenLabs Music v2](https://www.krea.ai/docs/api-reference/audio/elevenlabs-music-v2.md): Generate a song or instrumental track from a text prompt with ElevenLabs Music v2 — better prompt adherence, composition, and vocal delivery than v1, plus multilingual output. Vocals are included by default; set force_instrumental for instrumental-only output.
- [ElevenLabs Music v2.5](https://www.krea.ai/docs/api-reference/audio/elevenlabs-music-v25.md): Generate a song or instrumental track from a text prompt with ElevenLabs Music v2.5, the latest ElevenLabs music model. Vocals are included by default; set force_instrumental for instrumental-only output.

## OpenAPI Specs

- [openapi](https://api.krea.ai/openapi.json)