--- title: Key Capabilities at a Glance | Scenario Docs --- The Scenario API is a unified platform for AI-powered content creation — images, videos, 3D models, audio, and more — all accessible through a single REST API. This page gives you a quick overview of everything you can do. ## Table of Contents - [Image Generation](#image-generation) - [Video Generation](#video-generation) - [3D Model Generation](#3d-model-generation) - [Audio Generation](#audio-generation) - [Image Editing](#image-editing) - [Custom Model Training](#custom-model-training) - [Asset Tools](#asset-tools) - [Asset Management](#asset-management) - [How It All Works](#how-it-all-works) - [What’s Next?](#whats-next) --- ## Image Generation Generate images from text prompts using state-of-the-art models from Google, OpenAI, Meta, xAI, Ideogram, Recraft, BFL, and more. - **Custom-trained models**: Use your own trained LoRA to generate images in a specific style, character, or concept (see [Custom Model Training](#custom-model-training)). - **Text-to-image**: Describe what you want, get an image back. - **Image-to-image**: Pass a reference image (by asset ID) to guide generation with a text prompt. All image generation goes through the unified endpoint: `POST /generate/custom/{modelId}`. Browse available models: [Image Models](/get-started/generation/third-party-model-generation/index.md) --- ## Video Generation Create videos from text prompts, images, or existing videos using 24+ models from Google, OpenAI, Meta, Runway, Luma, Pika, Kling, MiniMax, xAI, HeyGen, and more. - **Text-to-video**: Generate a video from a text description. - **Image-to-video**: Animate a still image into a video. - **Video-to-video**: Transform an existing video with new styles or edits. Common parameters include `prompt`, `duration`, `fps`, `aspectRatio`, and `image` (asset ID for image-to-video). Each model supports different parameters — check the model’s page in the [Scenario web app](https://app.scenario.com) for details. Browse available models: [Video Models](/get-started/generation/video-generation/index.md) --- ## 3D Model Generation Convert images or text descriptions into 3D models using providers like Tripo AI, Meshy, Meta, Microsoft, Tencent, Deemos, and more. - **Image-to-3D**: Upload a reference image and generate a 3D asset. - **Text-to-3D**: Describe a 3D object and generate it from scratch. Configurable parameters vary by model and may include inference steps, guidance scale, target face count, and paint mode. Browse available models: [3D Models](/get-started/generation/3d-model-generation/index.md) --- ## Audio Generation Generate music, sound effects, and speech using models from ElevenLabs, Google, Meta, MiniMax, Beatoven, xAI, and more. - **Text-to-audio**: Generate audio from a text description or lyrics. - **Audio-guided generation**: Pass reference audio files (song, voice, or instrumental) to guide output. Parameters vary by model and may include `prompt`, audio file references, sample rate, and bitrate. Browse available models: [Audio Models](/get-started/generation/audio-generation/index.md) --- ## Image Editing Transform existing images using natural language instructions with image editing models like Google Gemini 3.1 Flash. - **Restyle**: Change the visual style of an image (e.g., “make this minimalist” or “convert to dark mode”). - **Modify**: Apply specific edits (e.g., “remove the text”, “change the background to a beach”). - **Refine**: Iterate on generated results with follow-up instructions. Image editing models use `referenceImages` (an array of asset IDs) to pass source images. See [Uploading Assets](/get-started/content/uploading-assets/index.md) for how to get asset IDs. --- ## Custom Model Training Train your own AI models (LoRAs) to capture a specific art style, character, or concept, then generate unlimited images in that style. - **Standard training**: Upload 5–15 example images to train a style or subject LoRA. - **Edit training**: Upload before/after image pairs to train an instruction-following editor. - **Multiple base models**: Train on FLUX.2 Dev, FLUX.2 Klein, Qwen Image, Z-Image, and more. Training workflow: 1. Create a model → `POST /models` 2. Upload training images → `POST /models/{modelId}/training-images` 3. Start training → `PUT /models/{modelId}/train` 4. Poll status until `trained` → `GET /models/{modelId}` 5. Generate with your model → `POST /generate/custom/{baseModelId}` with `modelId` in the body Learn more: [Training Custom Models](/get-started/training/training-models/index.md) --- ## Asset Tools Scenario provides utility models for common asset processing tasks — all accessible through the same `/generate/custom/{modelId}` endpoint. ### Background Removal Remove backgrounds from images automatically using models from Photoroom and Pixa. Browse available models: [Background Removal Models](/get-started/generation/background-removal-models/index.md) ### Image Upscale Enhance image resolution and quality using models from Magnific, Topaz Labs, Clarity AI, Recraft, BFL, and more. Browse available models: [Image Upscale Models](/get-started/generation/image-upscale-models/index.md) ### Video Upscale Enhance video resolution using models from Topaz Labs, Runway, Clarity AI, and more. Browse available models: [Video Upscale Models](/get-started/generation/video-upscale-models/index.md) ### Vectorization Convert raster images to scalable vector format using models from Recraft, Vision Cortex, and more. Browse available models: [Vectorization Models](/get-started/generation/vectorization-models/index.md) --- ## Asset Management Organize and manage all your generated content through the Assets and Collections APIs. - **Upload assets**: Upload images to get asset IDs for use in generation requests. See [Uploading Assets](/get-started/content/uploading-assets/index.md). - **Collections**: Group related assets (e.g., “Game Icons”, “Product Shots”) for easy organization. - **Captioning**: Auto-generate descriptions for your images using `POST /generate/caption`. --- ## How It All Works Every generation — image, video, 3D, audio, or tool — follows the same pattern: 1. **Send a request** to `POST /generate/custom/{modelId}` with your parameters. 2. **Get a job ID** back immediately (generation runs in the background). 3. **Poll for status** via `GET /jobs/{jobId}` until `succeeded` or `failed`. 4. **Retrieve assets** from the completed job’s `assets` array. Terminal window ``` # Example: generate an image curl -X POST https://api.cloud.scenario.com/v1/generate/custom/your-model-id \ -H "Authorization: Basic $(echo -n 'your-api-key:your-api-secret' | base64)" \ -H "Content-Type: application/json" \ -d '{"prompt": "a fantasy landscape, epic, detailed", "numOutputs": 2}' # Poll for results curl https://api.cloud.scenario.com/v1/jobs/{jobId} \ -H "Authorization: Basic $(echo -n 'your-api-key:your-api-secret' | base64)" ``` Use `?dryRun=true` on any generation request to preview the credit cost without actually generating. The available API parameters for each model are listed on the model’s page in the [Scenario web app](https://app.scenario.com). --- ## What’s Next? - **[Quick Start Guide](/get-started/documentation/quick-start-guide/index.md)** — Make your first API call in 5 minutes. - **[Integrating with Applications](/get-started/documentation/integrating-with-applications-real-world-scenarios/index.md)** — Real-world integration patterns with code examples. - **[Uploading Assets](/get-started/content/uploading-assets/index.md)** — How to upload images for use in generation. - **[Training Custom Models](/get-started/training/training-models/index.md)** — Train models on your own art style. - **[Workflows & Apps](/get-started/documentation/workflows-and-apps/index.md)** — Chain multiple generation steps into automated pipelines. - **[API Reference](https://cdn.cloud.scenario.com/static/api/swagger.yaml)** — Full OpenAPI specification.