AWS bedrock documentation change
Summary
Added comprehensive documentation for Luma Ray 2 model's text-to-video and image-to-video generation capabilities, including request examples, troubleshooting, performance notes, and related resources
Security assessment
The changes primarily add feature documentation and operational guidance. While troubleshooting mentions S3 permissions and HTTPS requirements, these are standard operational checks rather than addressing specific security vulnerabilities. No evidence of patching security flaws or responding to security incidents.
Diff
diff --git a/bedrock/latest/userguide/model-parameters-luma.md b/bedrock/latest/userguide/model-parameters-luma.md index 40c3c6b92..4edf85680 100644 --- a//bedrock/latest/userguide/model-parameters-luma.md +++ b//bedrock/latest/userguide/model-parameters-luma.md @@ -4,0 +5,2 @@ +Text-to-Video GenerationImage-to-Video GenerationTroubleshootingPerformance NotesRelated Documentation + @@ -70,0 +73,187 @@ The MP4 file will be stored in the Amazon S3 bucket as configured in the respons +## Text-to-Video Generation + +Generate videos from text prompts using the Luma Ray 2 model. The model supports various customization options including aspect ratio, duration, resolution, and looping. + +**Basic Text-to-Video Request** + + + { + "modelId": "luma.ray-v2:0", + "modelInput": { + "prompt": "an old lady laughing underwater, wearing a scuba diving suit" + }, + "outputDataConfig": { + "s3OutputDataConfig": { + "s3Uri": "s3://your-bucket-name" + } + } + } + +**Advanced Text-to-Video with Options** + + + { + "modelId": "luma.ray-v2:0", + "modelInput": { + "prompt": "an old lady laughing underwater, wearing a scuba diving suit", + "aspect_ratio": "16:9", + "loop": true, + "duration": "5s", + "resolution": "720p" + }, + "outputDataConfig": { + "s3OutputDataConfig": { + "s3Uri": "s3://your-bucket-name" + } + } + } + +**Additional Text-to-Video Example** + +Example with resolution and duration parameters. + + + { + "modelId": "luma.ray-v2:0", + "modelInput": { + "prompt": "a car", + "resolution": "720p", + "duration": "5s" + }, + "outputDataConfig": { + "s3OutputDataConfig": { + "s3Uri": "s3://your-bucket-name" + } + } + } + +## Image-to-Video Generation + +Transform static images into dynamic videos by providing keyframes. You can specify start frames, end frames, or both to control the video generation process. + +**Basic Image-to-Video with Start Frame** + + + { + "modelId": "luma.ray-v2:0", + "modelInput": { + "prompt": "A tiger walking in snow", + "keyframes": { + "frame0": { + "type": "image", + "source": { + "type": "base64", + "media_type": "image/jpeg", + "data": "iVBORw0KGgoAAAANSUhEUgAAAAgAAAAIAQMAAAD+wSzIAAAABlBMVEX///+/v7+jQ3" + } + } + } + }, + "outputDataConfig": { + "s3OutputDataConfig": { + "s3Uri": "s3://your-bucket-name" + } + } + } + +**Image-to-Video with Start and End Frames** + + + { + "modelId": "luma.ray-v2:0", + "modelInput": { + "prompt": "A tiger walking in snow", + "keyframes": { + "frame0": { + "type": "image", + "source": { + "type": "base64", + "media_type": "image/jpeg", + "data": "iVBORw0KGgoAAAANSUhEUgAAAAgAAAAIAQMAAAD+wSzIAAAABlBMVEX///+/v7+jQ3" + } + }, + "frame1": { + "type": "image", + "source": { + "type": "base64", + "media_type": "image/jpeg", + "data": "iVBORw0KGgoAAAANSUhEUgAAAAgAAAAIAQMAAAD+wSzIAAAABlBMVEX///+/v7+jQ3" + } + } + }, + "loop": false, + "aspect_ratio": "16:9" + }, + "outputDataConfig": { + "s3OutputDataConfig": { + "s3Uri": "s3://your-bucket-name" + } + } + } + +**Additional Parameters for Image-to-Video** + + * **keyframes** – (object) Define start (frame0) and/or end (frame1) keyframes + + * **frame0** – Starting keyframe image + + * **frame1** – Ending keyframe image + + * **type** – Must be "image" + + * **source** – Image source + + + + +## Troubleshooting + +Common issues and solutions when working with Luma AI models: + + * **Job Status "Failed"** \- Check that your S3 bucket has proper write permissions and the bucket exists in the same region as your Bedrock service. + + * **Image URL Access Errors** \- Ensure image URLs are publicly accessible and use HTTPS. Images must be in supported formats (JPEG, PNG). + + * **Invalid Parameter Errors** \- Verify aspect ratio values match supported options ("1:1", "16:9", "9:16", "4:3", "3:4", "21:9", "9:21") and duration is either "5s" or "9s". + + * **Timeout Issues** \- Use `GetAsyncInvoke` to check job status rather than waiting synchronously. Video generation can take several minutes. + + * **Prompt Length Errors** \- Keep prompts between 1-5000 characters. Longer prompts will be rejected. + + + + +## Performance Notes + +Important considerations for Luma AI model performance and limitations: + + * **Processing Time** \- Video generation typically takes 2-5 minutes for 5-second videos and 4-8 minutes for 9-second videos, depending on complexity. + + * **Image Requirements** \- Input images should be high quality with minimum resolution of 512x512 pixels. Maximum supported image size is 4096x4096 pixels. + + * **Output Video Size** \- Generated videos range from 5-50 MB depending on duration, resolution, and content complexity. + + * **Rate Limits** \- Async API calls are subject to service quotas. Monitor your usage and request quota increases if needed. + + * **S3 Storage** \- Ensure sufficient S3 storage capacity for output videos and consider lifecycle policies for cost optimization. + + + + +## Related Documentation + +For additional information and related services: + + * **Amazon S3 Configuration** \- [Creating S3 buckets](https://docs.aws.amazon.com/s3/latest/userguide/creating-buckets-s3.html) and [bucket policies](https://docs.aws.amazon.com/s3/latest/userguide/bucket-policies.html) for output storage. + + * **Async API Operations** \- [StartAsyncInvoke](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_StartAsyncInvoke.html), [GetAsyncInvoke](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_GetAsyncInvoke.html), and [ListAsyncInvokes](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_ListAsyncInvokes.html) API reference. + + * **Service Quotas** \- [Quotas for Amazon Bedrock](./quotas.html) for Bedrock service limits and quota increase requests. + + * **Video Processing Best Practices** \- [Submit prompts and generate responses with model inference](./inference.html) for general model inference guidance. + + * **Luma AI Documentation** \- [Luma Labs Video Generation Documentation](https://docs.lumalabs.ai/docs/video-generation) for detailed model capabilities and advanced features. + + + +